Canad. M. A. J.June 1, 1958, vol. 78 REVIEW ARTICLE 871

REVIEW ARTICLE

CORONARY VASCULARANASTOMOSES BY INTERNALMAMMARY ARTERYIMPLANTATION*

ARTHUR VINEBERG, M.D., C.M., Montreal

TODAY coronary atherosclerosis which results incoronary artery heart disease is a most contro-versial subject in regard to etiology, prognosis and,in particular, treatment. However, there seems tobe full agreement that narrowed or blocked coron-ary arteries may produce myocardial ischaemia.There is further agreement that anginal pain is areflection of the inability of myocardial fibresto obtain sufficient oxygenated blood to meetenergy demands. The end result of this complexbiochemical problem is a mechanical obstructionto the free flow of oxygenated blood into the myo-cardium.

QrtarIolcLr 3ones

normal

bring fresh extracardiac arterial blood into thenietwork. Proof of arteriolar connections betweenextracoronary blood sources and the myocardialnetwork is lacking. From the myocardial arteriolarnetwork (Fig. 2) arterioles communicate with (a)other myocardial arteriolar zones-collateral vessels;(b) lumina of ventricles-arterioluminal vessels;(c) myocardial sinusoids lying between myocardialfibre bundles-arteriosinusoidal vessels; (d) capil-

Tn-jocardtQl circuloittoi.

ischernic dcblused heart

Fig. 1.-Main coronary arteries branch to join separatearteriolar networks lying within the myocardium. In 10%of normal hearts these arteriolar zones are joined by inter-zonal arterioles. In diseased hearts these collateral arteriolesare almost always present (98%). In the normal or diseasedheart, vessels larger than capillaries have not been demon-strated entering the heart from extra-coronary sources.

Our present concept of the myocardial circula-tion in health and disease has been based uponcertain anatomical and pathological facts, a fewof which have either not been recognized or havebeen ignored by other workers. The first concerns

the extensive arteriofar network lying within themyocardium and receiving arterial blood directlyfrom the major coronary arteries. This network isdivided into separate arteriolar zones supplied bya branch of a major coronary artery. Normallythere is no communication between these zones,but the presence of ischaemias stimulates theformation of interzonal arterioles or collaterals(Fig. 1). These collateral arteriolar vessels joinarteriolar zones, thereby distributing blood more

freely throughout the heart muscle. Unfortunatelythese intracoronary anastomotic channels cannot

*Presented at the Annual Meeting of the International Col-lege of Angiology, 1957. From the Department of Experi-mental Surgery, McGill University; the Department ofCardiac Surgery, Royal Victoria Hospital; and the Instituteof Cardiology, Montr-al, Canada. This work has been madepossible through grants generously given by the Depa rt-ment of National Health and Welfare, Ottawa, Canada.

Fig. 2.-Schematic representation of normal myocardialcirculation. Zoned myocardial arteriolar networks are sup-plied by main coronary arteries. Prom these arteriolarplexuses, arteriole-sized vessels leave to communicate with(a) other myocardial arteriolar zones, (b) lumina ofventricles-arterioluminal vessels, (c) myocardial sinusoidslying between myocardia, (d) capillary bed about musclefibres, (e) venous system-arteriovenous fistula.

lary network lying about each myocardial fibre,and (e) the venous system.Coronary artery atherosclerosis generally in-

vades the major vessels in their epicardial course.The myocardial arteriolar network is usually freeof disease. When a coronary artery becomes nar-rowed, the volume of blood entering its arteriolarzone is diminished, thus decreasing the amount ofoxygenated blood available for distributionthroughout the network. This localized loweredoxygen tension stimulates the formation of col-lateral arterioles. These rapidly develop and,provided the neighbouring coronary arteries arehealthy, they carry blood from the non-ischixmicto the ischiemic zone. However, when disease hasaffected the supply coronary arteries of the neigh-bouring arteriolar zones, there is insufficient oxy-genated blood to spill over into the severelyischaemic zone. Once this situation has developed,vasodilators may relieve anginal pain but do notincrease the myocardial blood supply. Theoreti-cally, this can only be improved naturally by (1)the opening of arterioluminal vessels, (2) the en-

872 REVIEW ARTICLE Canad. M. A. J.June 1, 1958, vol. 78

There is much confused thinking at present withregard to anastomoses between the coronary cir-culation and the extracoronary blood supply. Un-questionably, anastomoses exist at the points ofpericardial reflection over the great vessels as theyenter and leave the heart. Unfortunately, thesemuch described anastomoses are capillary in size,are superficial, and are mostly with the vessels ofthe atria. There is no evidence at present to showthat these existing capillary anastomoses enlarge inresponse to myocardial ischaemia or followingbilateral internal mammary artery ligation.

If the above concept of the coronary circulationin health and disease is correct, surgical procedures

Fig. 3.-Drawing showing internal mammary artery de-tached from chest wall and buiried in myocardial tunnelmade in left ventricle.

largement of thebesian canals, (3) the develop-ment of arteriole-sized communications betweenthe myocardial arteriolar network and an extra-coronary blood source. There is little anatomicalevidence in normal or diseased hearts to show thatany one of these three sources is ever tapped. Thechance of myocardial survival with all maincoronary vessels blocked is very slim indeed. Theoccasional case of this type previously reportedmust be re-examined in the light of more recentmass injection studies by Schlesinger. Perhaps aconus branch or other aberrant coronary vessel wasresponsible for myocardial survival in the reportedcases.

5cKLe5situ,,era 5ohtti.or"-rovxredre erecl

:. /Iaq

irtjzqactio-m ot

/soVCtiot\\OI

xiti

_circunmdlerx a.

tn-irpxodri .irmttrrla

x%ammrxmrg a.

Fig. 5.-Drawingimplanted internalleft ventricle andcoronary artery.

to show Schlesinger mass injected intomammary artery, fllling arterioles ofcoming out of cannula placed in left

are needed to relieve the myocardial ischemiacaused by the blocking of the main coronaryarteries by atherosclerosis.There are various possible surgical approaches

to the problem:1. To *resect the diseased coronary artery and

either anastomose a systemic artery to its distalend or insert a homograft. Experimentally at least50% of these grafts block, which is the experiencewith replacement grafts in small vessels in thehuman extremities.

2. To scrape out the obstruction by endarterec-tomy. This may be of value in some cases but,again, following the experience of endarterectomyin other areas, it may prove not too successful.

3. To obtain wider distribution of blood bycoronary sinus ligation (Beck I). Very little extra-coronary oxygenated blood is delivered after thisoperation.

4. To perform a by-pass operation, in which thepoints of coronary artery obstruction are by-passedby:

(a) application of surface grafts which formsuperficial coronary vascular anastomoses,, or

(b) vascular myocardial implants which, the-Fig. 4.-Dissected left subelavian artery and internal oretically, are capable of pouring fresh oxygenated

mammary artery, showing internal mammary artery in bL t th* 1_ *r*r* r *s1left ventricle seven months after implantation. blood into the entire myocardial arterioiar system.

--P.

IJanad. M. A. J. R I T EJune 1, 1958, vol. 78 REVIEw ARTICLE 873

The vascular implant operative procedures are I M h 1 Jdesigned to pour extracardiac blood into the ven- _latricular arteriolar network at points distal to theareas of coronary artery disease.Twelve years ago the left internal mammary

artery was detached from the chest wall and buriedwithin a tunnel in the left ventricle of an animal'sheart (Figs. 3 and 4).To the astonishment of all, the buried artery

not only remained open, but formed a connectionbetween itself and the coronary circulation (Fig.5). During the ensuing years well over 1500 ani-mal experiments have been performed in orderto determine why an artery, tied and cut at its

br tl%e CCLr&Uz_-i,6.c^wa.WoC Entwnt Mamua~- rtirM_*je *

Fig. 7.-Digestion cast of animal heart made to showImplanted internal mammary artery lying in myocardium.Part of cast has been cut away, revealing early branchingof internal mammary artery 12 days after implantation.

8s ar~~~OpUlar b FtV"A -

tAod erv.A.--ovib fo;

t4;6vtar 'a7 true arterioles (Fig. 7). These arterioles within afew weeks join directly with the arterioles of theventricular myocardium, thus forming mammary-coronary anastomoses through which extracardiac

__ . arterial blood is delivered to the myocardium, and_dfl sj~_ _ _ _ which are still present many months to years later

(Fig. 8).It has been found that the frequency and dura-

tion of mammary-coronary anastomosis is depen-(ntnct 0{rt) W~~~~~den Bupon:

Fig. 6.-Drawing of implanted internal mammary artery dent upon:in left ventricle with opening in side of vessel. Blood escapes 1. The technique of internal mammary arteryfrom Internal mammary artery into myocardial sinusoids,which is why the implanted vessel remains open until its preparation and implanatation, andown branches Join the coronary arterioles.

distal end, and buried in a new environment, suchas the myocardium, remains open for years. Ourinvestigations showed that the artery remains openif it is allowed to bleed from an open side branchinto the myocardial tunnel where the escapedblood is taken away through the medium of themyocardial sinusoids (Fig. 6). After 12 days theimplanted internal mammary artery starts to bud

TABLE I.-ORGAN VASCULAiIZATION BY ARTERY OR PROSTHESISIMPLANTATION

Author Artery or prosthesis Organ

Experimental:Vineberg, Montreal (a) Internal mammary

artery .............. Left ventricle(b) Double aortico-homo-

graft.. Left ventricle(c) Polyethylene thebesian

canal .........Left ventricleGlenn, New York Internal mammary artery Left ventricleLiebow, New Haven Splenic artery.. Left ventricleBlalock, Baltimore (a) Internal carotid. Left ventricle

(b) Aortico-homograft. Left ventricleWilson,

Bristol, England Internal mammary artery Left ventricleDavis, Miami, Florida Splenic artery . ........ Kidney and liverSmith,

Bradington, Florida (a) Aortico-homograft..... Left ventricle(b) Aortico nylon tubeL.eft ventricle

Massimo,Florence, Italy Polyethylene thebesian

canal . Left ventricle Fig 8.-Digestion cast of animal heart made by injectingHuman coronary artery disease: implanted internal mammary artery with vinyl plastic57 cases, seven months after implantation. Part of cast cut away to.

Vineberg, Montreal Internal mammary artery.. Left ventricle show size of branches which have grown from implanted'35 cases, internal mammary artery and have joined with surround-'Walker, Charleston Internal mammary artery Left ventricle ing coronary arterioles.

874 REVIEW ARTICLE Canad. M. A. J.June 1, 1958, vol. 78

I implant into the ventricle. Smith has.. | l-found that aortico-homografts as well

as aortico nylon tube grafts remain_ 0 - -_ open and form anastomoses. In our

hands this has occurred but the per-centage of thrombosed grafts wasextremely high. Davis'4 (Fig. 9) hasfound that vascular anastomoses de-velop between an implanted splenicartery and the vessels of the kidneyand liver after implantation.

In 1950, in our laboratory, an at-tempt was made to create artificialthebesian canals by placing an open

Fig. 9.-(a) Roentgenogram showing splenic artery implanted into kidney. polyethylene tube or homograft inOutline of kidney circulation inade by injecting implanted splenic arterywith radiopaque medium. (b) Roentgenogram showing splenic artery im- the left ventricular lumen and bury-planted into liver. Outline of liver circulation nmade by injecting radiopaquemedia through implanted internal miiammary artery. (Photograph by courtesy ing the other end in the myocardium.of Davis.)

2. The character of the myocardium into which 5the implant is placed.'-'0The idea that an internal mammary artery could

be detached from the chest wall and left in amyocardial tunnel without immediate or delayed R

Fig. 12.-Photograph of ameroid constrictors used to con-strict slowly coronary arteries at their origins in animals.Note ameroid casein plastic with slot to fit over artery andencased in steel jacket. The casein plastic absorbs water,swells, and constricts coronary artery. Rate of water absorp-tion controlled by coating ameroid surface with Vaseline.

Some of these remained open (Fig. 10) but seldomshowed vascular communications with the ventri-cular arteriolar network, and therefore the pro-

:. . Not~~~~~~~edureriwaseiab andoned. lttoft vr rer n

More recently, using the principles upon whichinternal mammary artery implantation has beenbased, Massimoe'6 has succeeded in obtaining avascular anastomosis between the left ventricular

Fig. 10.-Artificial thebesian canal made by insertingpolyethylene tube into left ventricular lumen and buryingother end in myocardium. Note patent lumien. (Reportedby author in this Journal 1953.)

thrombosis has been for a long time rejected. How-ever, that fact has been confirmed by many workers(Table I), and the principles upon which it isbased utilized for other vessels in the heart, kidneyand liver. Kline" reports that the splenic, and bfBlalock10 the carotid artery, remains open after

Fig. 11.-Drawing of polyethylene T-tube lying in myo-cardium and connecting with left ventricular lumen. (Cour-tesy of Massimo, J. Thoracic Sturg.)

Fig 13.-Photograph to show ameroid constrictors onanterior descending and circumflex arteries.

ovk,.,,.,m ,e;,e,,

Canad. M. A. J.June 1, 1958, vol. 78

I(a) (J) {C)Fig. 14.-Photomicrograph of sections through anterior descending artery (a) proximal,

(b) through, (c) distal to site of ameroid constriction. Note narrowing of coronary arteryproximal to and at site of ameroid constrictor compared with normal artery lumen just distalto point of arterial constriction.

cavity and the arteriolar sinusoidal myocardialmeshwork (Fig. 11).Now, there is little doubt that an implanted in-

ternal mammary artery sends out branches whichjoin with the arteriolar ventricular network. Theexperimental proof of the value of this vessel andthe frequency of mammary-coronary anastomosishas awaited a satisfactory method of producingchronic coronary artery occlusion which, fortun-ately, was developed in our laboratory two yearsago by Litwack.An ameroid tube encased in a steel jacket is

placed around the origins of the anterior descend-ing and cir-cumflex coronary arteries (Figs. 12and 13). The ameroid material absorbs water,expands, and slowly constricts the two coronaryvessels. The rate of water absorption may be variedby coating the ameroid with various substances.A coating of Vaseline (petrolatum) prolongs thetime taken for the ameroid to absorb water, swell,and constrict the coronary vessels. The thicknessof the Vaseline coat varies according to the lengthof time the ameroids are left in the Vaseline bath

and the temperature of the bath. Ameroids coveredwith a thin coating of Vaseline and placed aroundthe anterior descending and circumflex branches ofthe left coronary artery caused death of all con-trol animals within 28 days, whereas those moreheavily coated took an average of 14 weeks tokill the control animals. In a recent series of controlanimals, ameroid constrictors were placed upon theanterior descending and circumflex arteries byDeschene, and all controls were dead within 14weeks. Death usually occurred suddenly duringexercise or excitement while at the animal farm.Four of the animals of this series, brought backto the laboratory, dropped dead after being de-liberately excited; in each instance, when the thoraxwas opened, there was ventricular fibrillation. Thedegree of arterial constriction amounted to a re-duction of internal diameter varying between one-half and one-third of normal (Fig. 14).

In a series of 20 internal mammary artery im-plantations carried out at the same time as theplacing of ameroid constrictors (Table II), theanimals which were exercised or excited by dog

TABLE II.-VALUE OF INTERNAL MAMMARY ARTERY IMPLANTATION IN PROTECTING THE VENTRICULAR MYOCARDIUMMADE ISCH.EMIC BY CORONARY AMEROID CONSTRICTORS

Time after internal Cause of deathmammary arteryimplantation Exercise and excitement

+ Extent and frequencyameroid coronary Number of of mammary-coronary Induced

constrictors animals anastomoses Natural Moderate Severe Sacrificed2 to 4 weeks5 to 12 weeks

20 to 30 weeks

586

Anterior ventricular wall........ 3 of 5 5Anterolateral ventricular wall.... 8 of 8 8Entire left ventricle ............. 6 of 6 0 0 0

(Unable to kill by exercise)Total open internal mammary artery implants-17 of 19=89%

Mammary-coronary anastomoses-16 of 19=84%-NOTE: Of 20 controls 17 died within 29 weeks naturally (average 14 weeks)

3 died following induced moderate exercise.

REVIEW ARTICLE 875

5

876 REVIEW ARTICLE Canad. M. A. J.June 1, 1958, vol. 78

Fig. 15.-Roentgenogram of left ventricle injected withSchlesinger mass through implanted internal mammaryartery less than five weeks after imnplantation. Animaldied because of myocardial isch.enia caused by constriction ofanterior descending and circumnflex arteries by ameroidconstrictors. Note: At this stage internal mammary arterycommunicated with arterioles of the anterior ventricularwall only.

fights less than five months after implantation died,and yet there were patent internal mammaryarteries with mammary-coronary anastomoses.However, injection of these internal mammaryarteries with Schlesinger's mass showed that underfive weeks the internal mammary artery anasto-mosed only with the arterioles of the anterior ven-tricular wall (Fig. 15). Between five and 12 weeksthey extended to include the antero-lateral leftventricular arterioles, and the injection mass didnot reach the posterior left ventricle wall. Therewere six animals which lived beyond five monthsand could not be killed by hard exercise. One diedlater from distemper. Five were sacrificed and in

Fig. 16 (a).-Roentgenogram showing circumferentialfilling of left ventricle by Schlesinger mass injected throughInternal mammary artery implanted more than five monthspreviously. Note: This animal could not be killed by severeexercise, even though anterior descending and circumflexarteries were markedly narrowed at their origins by ameroidconstrictors.

every instance the Schlesinger mass injectedthrough the internal mammary artery had filled thearterioles of the entire left ventricle (Fig. 16).

In the series of ischaemic animal hearts, 89% ofthe implanted internal mammary arteries werepatent and 84% formed mammary-coronary anas-tomoses. There was minimal to no intimal proli-feration in the internal mammary arteries left insuch ischtemic myocardia.

SELECTION OF HUMAN PATIENTS WITHCORONARY ARTERY HEART DISEASE FORREMEDIAL SURGERYIn treatment of human coronary artery disease,

selection of the type of patient for remedial surgeryis very important. In 1950, when the first internalmammary artery implant was performed upon ahuman case, there was no opportunity to selectpatients. The only patients reporting to theMontreal group for surgery were desperate, dis-abled, and pain-ridden people-a large percentageof whom were bed-chair invalids. Experiencequickly showed that the proper selection of patientsfor coronary heart disease surgery should belargely dependent upon the stage of developmentof their disease, and the following criteria wereagreed upon:

1. Proven coronary artery disease, with typicalangina pectoris pain. The presence of coronaryartery disease in itself is not sufficient for surgery,unless the patient suffers anginal pain which istruly causing difficulty.

2. At least one to two years of adequate medicaltreatment.

Fig. 16 (b).-Same heart sectioned to show full thicknessof left ventricular wall.

Canad M. A. J.June 1, 1958, vol. 78 REVIEW ARTICLE 877

CONTRAINDICATIONSThere are certain contraindications to internal

mammary artery implantation; for example, surgeryshould NOT be performed

(a) When the patient is asymptomatic.(b) When there is evidence of disease activity.(c) When there is evidence of left ventricular

failure.(d) When other incurable disease co-exists.(e) When the patient has angina decubitus.

EVALUATION OF RESULTSEvaluation of the treatment of coronary artery

disease is difficult because of the numerous andunpredictable variations which occur during thenatural course of the disease, and because at thepresent time there is no objective laboratorymethod of measuring improvement in myocardialcirculation. At present, the observer must rely uponthe following criteria as evidence of improvedmyocardial circulation, namely:

(a) Survival.(b) Disappearance of anginal pain for more

than six months.(c) Increase in work and exercise tolerance.(d) Improvement in electrocardiogram.(e) Failure to develop fresh infarction.In our experience there are two main groups of

patients in whom entirely different results may beexpected from revascularization surgery. Theseare:

1. Cases with angina at rest, without excitingcause (angina decubitus).

2. Cases with no angina at rest, without excitingcause. In previous publications the first group havebeen referred to as cases of "angina decubitus".This term, which refers to the onset of anginalpain when the patient assumes the decubitusposition, has led to much misunderstanding anddoes not satisfactorily describe this group ofpatients. In this group, patients while at rest ex-

perience anginal pain for which there is no obviousexciting cause, such as talking, telephoning, watch-ing television, or emotion. The coronary arterycirculation is so reduced that the basic myocardialmuscle energy demand necessary to maintain lifeat rest cannot be completely met, and myocardialischaimia and anginal pain develop while at com-

plete rest. Such hearts are beyond help since toomuch of the myocardium has been destroyed.

In the second group of patients, with no anginaat rest without exciting cause, the situation isdifferent. Many of the patients in this group havepain at rest but always associated with an excitingcause, such as eating. Many are awakened by painduring sleep, under which circumstances the ex-

citing cause appears to be a drop in blood pressureor a disturbing dream. All patients in this groupwere able to walk at least 50 feet without pain andmost were disabled and unable to work. Pain at

rest, when it occurred, could always be related tosome exciting cause.

OPERATIVE MORTALITYThe mortality for any operative procedure de-

pends upon: (i) the method of estimating operativemortality, and (ii) the selection of patients.

In this series of patients, any death of a personwithin 28 days of operation, in or out of hos-pital, is listed as an operative mortality. In coronaryartery heart disease, postoperative deaths aremostly due to a fresh coronary artery thrombosis.This may occur during operation, immediatelyafter operation or, as in one of our patients, threeweeks after operation. It is quite true that thepatient died of his disease, but it was still anoperative death and cannot be excluded fromoperative mortality statistics as has been suggestedby Glover.18 Dana and Ohler19 have found thatthe mortality of patients suffering from anginapectoris at the time of undergoing general surgeryis 7.8%.With regard to case selection, it is obvious that

the operative mortality rate should scale downfrom the far advanced cases to those which areasymptomatic.

In our series of 59 patients the cases have beendivided into those with angina at rest withoutexciting cause-17 cases, with an operativemortality of 59%o, and those with no angina at restwithout exciting cause-40 cases, operative mortal-ity 5% (Table III).

TABLE III.-INTERNAL MAMMARY ARTERY IMPLANTATIONOPERATIVE MORTALITY WITHIN 30 DAYS OF OPERATION

CasesAngina at rest without exciting

cause .................... 17 10 deaths.... 59%No angina at rest without excit-

ing cause ................. 40 2 deaths.... 5%Total number of cases.... 57

TABLE IV. INTERNAL MAMMARY ARTERY IMPLANTATIONLATE MORTALITY

4 MONTHS TO 6 YEARS AFTER OPERATION

Angina at rest without exciting cause:Total number of cases ........................... 17Survived operation ............................. 7

3 died 12 months to 4 years after operation:(i) Cancer of pancreas.(ii) Aplastic anemia.

(iii) Heart.(2 patients pain-free and working.)No angina at rest without exciting cause:

Total number of cases .......................... 40Survived operation ............. ................ 38

8 died 4 months to 6 years after operation:(i) Rupture right ventricular aneurysm.(ii) Dissecting aortic aneurysm.

(iii) Died pushing car out of snow-rupture ventricular scar.(iv) Died after riding horseback.(v and vi) Right ventricular infarction.(vii andc viii) Known cardiac deaths.(5 patients had no pain or slight pain and were working at time

of death).

878 REVIEW ARTICLE Canad. M. A. J.June 1, 1958, vol. 78

TABLE V.-RESULTS OF INTERNAL MAMMARY ARTERY IMPLANTATIONIN ANGINA AT REST WITHOUT EXCITING CAUSE

..... 17 OPERATIVE DEATHS...... 10 (57%) . LALTE DEATHS...... 3 (18%)

No pain or slight pain................................

Less pain...........................................Same or worse......................................Returned to work....................................No. original group still alive...........................

Working full time....................................No pain or slight pain................................Less pain...........................................Same or worse......................................

LATE MORTALITYThe cause in late deaths of patients of bot

groups is shown in Table IV. Three were noi

cardiac deaths, and two were caused by riglventricular lesions.

POSTOPERATIVE RESULTSAn analysis of the patients with angina at re

without exciting cause reveals that, in addition ta high mortality rate in this group, only or:patient is working full time, and only two patienor 12% of the original group may be considereto have improved after internal mammary arteximplantation (Table V), whereas, for the 4patients with no angina at rest without excitincause, although 80% were unable to work befoioperation, 78%o of the survivors and 70% of tIoriginal group were pain-free, or had slight paior less pain, and 88% of the survivors returned Iwork.A recent survey made six months to seven yea

after operation reveals that 30 patients, or 75% (

the original group, are still alive. Twenty-four,60%o, are working full time-16 of these 3 years t6 years 9 months after operation, and 8 of thei6 months to 2 years 9 months after operatioiTwenty-five or 66% are still improved, being pailfree or having slight pain or less pain (Table VI

3 (43%)86% of survivors improved.35% of original group

3 (43%)1

4 (57% of survivors-24% of original group)4 (24%) 2 (3 years 9 months to 6 years 11 months)

2 (2 years 11 months)1 (6% of original group)1 (6%)1 (6%)1

During the past 15 years much criticism hasth been levelled at the internal mammary artery

implant procedure. Perhaps the most serious ofnt these criticisms has been the statement by others

that the implanted arteries block, by thrombosisor intimal proliferation, when left in animal heartsfor more than five months. However, the authorsof these statements have found that, when our

st technique has been carefully repeated, theto ischoemic hearts of animals are successfully re-ie vascularized just as we have shown early in thists paper. Our statements based on animal experi-d mentation have now received confirmation inry pathological studies made on the implanted in-t0 ternal mammary arteries of 14 patients who diedig from six hours to four years after operation.re

EARLY DEATHS[n

to Of the 14 patients studied, 8 died within 21 daysof operation. Seven of these had had a fresh coron-

rs ary occlusion and the internal mammary artery wasof open. In two cases injection of the implant was

Dr done with India ink which entered the coronaryto vessels 82 hours to 21 days after operation.

n. LATE DEATHS

n- A study of the six patients who died four). months to four years after operation showed that

TABLE VI.-RESULTS OF INTERNAL MAMMARY ARTERY IMPLANTATIONIN No ANGINA AT REST WITHOUT EXCITING CAUSE

TOTAL NUMBER OF CASES...... 40 OPERATIVE DEATHS...... 2 (5%) LATE DEATHS...... 8 (20%)

Unable to work .............. 32 (80%)

Condition of 38 survivors (95%)-6 months to 7 years after operationNo pain or slight pain ................................ 27 (75%/,)

78% of survivors70% of original group improved

Less pain ............................................ 1 ( 3%h)Same or worse ....................................... 8 (22%)Returned to work .................................... 32 (88% of survivors) (82% of original group)Number of original group still alive ...................... 30 (75%)Working full time .................................... 24 (60%) 16-3 years to 6 years 9 months.

8-6 months to 2 years 9 months.

No pain or slight pain ................................ 24

25 (66%) still improvedLess pain ............................................. 1Same -or worse..........5

(2 cases less than 6 months after operation not included)

TOTAL NUMBER OF CASES.

.Condition of 7 survivors

Unable to work before operation .17 (100%)6 months to 7 years after operation

Canad. M. A. J. REVIEw ARTicLE 879June 1, 1958, vol. 78

TABLE VII.-CONDITION OF HUMAN INTERNAL MAMMARY ARTERY6 HOURS TO 4 YEARS AFTER IMPLANTATION

Operativedeaths Number T'ime after Open

autopsied cases implantation Cause of death 1.M1. artery

8 2 6 to 24 hours (a) Fresh thrombus left coronary. 2(b) Fresh thrombus circumflex

2 60 to 24 hours Fresh thrombus circumflex .. 21 82 hours Penicillin allergy .. 11 4 days Fresh thrombus right coronary . . 11 10 days Fresh thrombus right coronary . . 01 21 days Fresh thrombus right coronary . . 1

8 7(88%)Late deathsautopsied

6 1 4 months Right coronary artery occlusion .. 02 18 months (a) Dissecting aortic aneurysm. 1

(b) Cancer of the pancreas .11 3 years Rupture right ventricle ..

1 3 years 5 months Cardiac irregularity .. 11 4 years Aplastic anwemia .. 0

6 4(67%)

Total cases. .14 11(78.5%)

the internal mammary artery was open in fourof the six cases. In one patient, who died of aruptured right ventricular aneurysm, the onlyartery open in the heart three years ofter operationwas the implanted internal mammary arteryplaced there three years before. The implantedvessel had a 60 to 70%o lumen, and there appearedto have been no progress of the disease in theleft ventricle from the time of operation to thetime of death (Table VII).One of the most interesting cases is that of a

60-year-old man who, at the time of internal mam-mary artery implant, was completely crippled withsevere angina, and who returned to an 8-hournight shift in an aircraft factory where he workedfor 18 months, pain-free, until he died of cancerof the pancreas. His heart was returned to Montreal,

..... .:. .. ....... 3~~~~~~~~~~~~~~~~~~~~Fig. 17. Photomicrograph of internal mammary artery

one-third along the tunnel. Note active branching. Thisphotomicrograph was made from the heart of a 60-year-oldpatient with angina at rest without exciting cause. Patienttotally disabled before implantation; returned to eight-hourdaily work, pain free; died of carcinoma of pancreas. Thecircumflex artery was injected antegrade with India inkand the injection mass w as found in the internal mammaryartery.

where 850 serial sections were made of the im-planted internal mammary artery. It was found thatthe implanted internal mammary artery was widelyopen, without intimal proliferation at the entranceto the tunnel (Fig. 17).Another patient who showed marked clinical

improvement following internal mammary arteryimplantation returned to work, and died 3 years5 months after operation. The right and left coron-ary arteries were cannulated and the internal mam-mary artery was injected with barium mass undergravity pressure at the Ford Hospital in Detroit byDr. Edward Priest, who reported that the injectionmass entered the internal mammary artery, fillingthe arterioles of the entire left ventricle, and sub-sequently appeared in the cannula which wasplaced in the left coronary artery.

REFERENCES

1. VINEBERG, A. M.: Canad. M. A. J., 55: 117, 1946.2. VINEBERG, A. M. AND JEWETT, B. L.: Ibid., 56: 609, 1947.3. VINEBERG, A. AND MILLER, G.: Ibid., 64: 204, 1951.4. 'VINEBERG, A. M. AND NILOFF, P. H.: Surg., Gyntec. &

Obst., 91: 551, 1950.5. BENCOSME, S. A. AND VINEBERG, A.: Am. Heart J., 45:

571, 1953.6. VINEBERG, A. M.: J. Thoracic Surg., 18: 839, 1949.7. VINEBERG, A. AND MILLER, D.: Am. Heart J., 45: 873,

1953.8. VINEBERG, A. et al.: J. Thoracic Surg., 29: 1, 1955.9. VINEBERG, A. AND BULLER, W.: Ibid., 30: 411, 1955.

10. GLENN, F., HO,LSWADE, G. R. AND GORE, A. L.: The fateof an artery implanted in the myocaldium; In: Sur-gical Forum Clinical Congress of the AmericanCollege of Surgeons, 1950, W. B. Saunders Company,Philadelphia, 1951, p. 289.

11. KLINE, J. L. et al.: Am. J. Path., 32: 663, 1956.12. BLALOCK, ALFRED: Reported at Southern Surgical Asso-

ciation Meeting, Boca Raton, Florida, December 1956.13. WILSON, MICHAEL G.: University of Bristol (England)

Department of Surgery-Personal communication.14. DAVIS, H. CLINTON: University of Miami, Florida-Per-

sonal communication.15. SMITH, S. et al.: Surg., Gynec. & Ob8t., 104: 3, 263, 1957.16. MASSIMO, C. AND BOFFI, L.: J. Thoracic Surg., 34: 257,

1957.17. VINEBERG, A. AND WALKER, J.: Amt. Heart J., 54: 851,

1957.18. GLOVER, R. P. et al.: J. Thoracic Surg., 34: 661, 1957.19. DANA, J. B. AND OHLER, R. L.: J. A. M. A., 162: 878, 1956.