The Discovery of Dicumarol andwilliams.medicine.wisc.edu/warfarin_discovery.pdf · 2011-11-25 · 98INKL headed the Biochemistry Department of the University of Minnesota. Hehad offered

The Discovery of Dicumarol and Its SequelsBy KARL PAUL LINK, PH.D.

THE STORY of Dieumarol* has been toldseveral times by me in the past 17 years,

and often by others. Like any good story itneed not be told in exactly the same manner

each time. In Wisconsin it has become a

kind of legend. I shall consider only the highwater marks of certain chapters.

Fortunately the basic scientific facts on thediscovery and development have already beenthoroughly recorded1-6 so that little new in-formation on Dieumarol and its sequels needsto be revealed here. However, when I dointroduce new material it will be restricted tothat which is documented or sustainable viaiemoranda or letters.The story begins some 36 years ago on the

prairies of North Dakota and in Alberta, Can-ada. In the 1920's a new malady of eattle in-volving fatal bleeding showed up almost si-multaneously in these areas. The veterinarians,Schofield and Roderick, were forced to con-elude that the cause of the disease was neithera pathogenic organism nor a nutritional de-ficiency. The origin of the new malady was

traced to stacks of sweet clover hay mysteri-ously gone bad. Hence the disease becameknown in veterinary practice as "sweet cloverdisease" and it was found that it was causedonly by improperly cured hay made from thecommon varieties of sweet clover. When firstobserved this disease was in a sense withoutparallel in animal pathology or human medi-cine. When cattle or sheep ate the spoiledhay the disease slowly became manifest by a

progressive diminution in the clotting power

of the blood (about 15 days) and resultantinternal hemorrhage which usually becamefatal in about 30 to 50 days.

From the Department of Biochemistry, Universityof Wisconsin, Madison, Wis.

Given February 15, 1958 at the New York Academnyof Medicine under the auspices of the Section of Medi-

cine and the Ne-wv York Heart Association, oil the

programme, "'The Historieal and Physiological As-

pects of Ainticoagulanelts.'"

97

It was. recognized by Schofield and Roder-ick that the disease was reversible. It couldbe controlled in cattle by the withdrawal ofthe spoiled hay from the diet and by transfu-sion of blood freshly drawn from normal cat-tle, provided the hemorrhagic extravasationhad not proceeded too far. Indeed, theyshowed that even in desperate cases, recoverycould be hopefully anticipated after trans-fusion and change in diet (good hay).

In a comprehensive and thorough study ofthe pathology and physiology of the diseaseRoderick in 1931 emphasized that the delayedor abolished coagulability of the blood wasdue to a "prothrombin" deficit. Indeed heshowed that the severity of the hemorrhagiccondition paralleled the reduction in "pro-thrombin content or activity." He did thisby using the technic developed by that greatAmerican pioneer of blood coagulation, thelate Professor W. H. Howell. Solutions ofwhat was then called "prothrombin" pre-pared by precipitation of normal bovine plas-ma with Howell's acetone method when addedto the "sweet clover blood" promoted coagu-lation. In contrast, preparations of "pro-thrombin" made in a parallel manner from"sweet clover blood" did not produce coag-ulation. The other constituents for the main-tenance of normal coagulability known at thattime (fibrinogen, calcium, platelets, andinhibitory substance) appeared to be un-affected.

I first learned about the hemorrhagic sweetclover disease of cattle in December 1932through the late Ross A. Gortner, who then

*Djicumarol is the trademark for 3,3'-methylenebis-(4-hydroxycoumarin). The anticoagulant was madeavailable in 1940 and 1941 for clinical use by thecooperative efforts of the Wisconsin Alumni ResearchFoundation, Madison, Wis., the Abbott Laboratories,North Chicago, Ill., Eli Lilly and Company, Indiana-polis, Ind., and E. R. Squibb and Company, NewBrunswick, N. J. The official U.S.P. name is bishy-droxycoumarin.

Circulation. Volwme XIX, January, 1959

at UNIV OF WISCONSIN MADISON on October 4, 2011http://circ.ahajournals.org/Downloaded from

http://circ.ahajournals.org/

98INKL

headed the Biochemistry Department of theUniversity of Minnesota. He had offered me

a post and I had come to St. Paul to considerit. Since the "sweet clover disease" was alsoa problem in Minnesota it was one of the pro-

jects open for study if I chose to accept. Itwas Gortner who supplied me with the origi-nal publications of Roderick. Some attemptshad been made in Gortner's department toextract the hemorrhagic agent but they, likethose of Roderick and others, had failed.

Curiously, the "official start of our workin January 1933 in cooperation with ProfessorR. A. Brink and W. K. Smith of our Geneticsdepartment was on a different aspect of thesweet clover problem. They sought to developa strain of sweet clover suitable for Wisconsinclimatic conditions low in, or free from, cou-marin. Though coumarin smells sweet (thecharacteristic smell of new mown hay is due toits presence) it tastes bitter, and it was knownthat the -bitter taste of green sweet cloverplants, Melilotus alba and M. officianalis,paralleled the total coumarin content. In ac-

tual practice it was observable that cattle (orrabbits) would eat the less bitter plants first.

Tragedy out on the Farm

Quite apart from the "official" start con-

cerned primarily with the palatability ques-

tion my laboratory had a direct catalytic hitfrom agricultural practice.

Indeed on a Saturday afternoon in Febru-ary 1933 following the first conferences withBrink, while a blizzard was howling and themercury was hovering near zero, a farmerfrom the vicinity of Deer Park, Wisconsin,some 190 miles from Madison appeared withwhat the late Professor A. J. Carlson mighthave called "the evidence." Curiously thefarmer's name was Ed Carlson. The hemo--rhagic sweet clover disease of cattle was ramp-

ant on his farm. He had fed sweet clover hayfor years previously without encountering anydifficulties and he doubted the veterinarian'sdiagnosis. Accordingly he was advised to go

to the Agricultural Experiment Station au-

thorities to get the facts. The office of theState Veterinarian had closed and pure chancehad brought him to the Biochemistry Building.

Farmer Carlson's multiple evidence was adead heifer, a milk can containing blood com-pletely destitute of clotting capacity, andabout 100 pounds of spoiled sweet clover-theonly hay he had to feed his cattle.

His account of the over-all course of thedisease coincided perfectly with the classical" sweet clover poisoning" picture. Late inDecember he had lost 2 young heifers. In Jan-uary 1 of his favorite old cows had developeda massive hematoma on a thigh and followinga skin puncture fatal bleeding set in rapidly.Finally 2 young cows had died on Friday andthe bull was oozing blood from the nose. Sohe took off for Madison in a blizzard.

I immediately had to tell farmer Carlsonthat we could do no more at this time than torecommend the teachings of Roderick andSchofield. He had to stop feeding that hay,and possibly transfuse those desperately sickcattle, if he wanted to save them. Eventuallyit might become possible to make some usablerecommendations to avoid such disasters, butnot now.

I can still see him take off for home about4:00 p.m. Those 190 miles of drifted roads be-tween our laboratory and his barn must haveappeared to him like a treacherous and som-ber ocean.

I cannot take the time to tell all the detailsof this slice of the Dicumarol story, but Ican assure you its impact on me was immense.I will relate a part of it exactly as I did inmy first lecture on Dicumarol given at theMayo Clinic on March 12, 1942.When farmer Carlson came to see us, my

senior student and old man Friday was EugenWilhelm Schoeffel, a volatile Schwabian whocame to the U.S. in 1926 with a diploma inAgricultural Chemistry. After serving a 2-year apprenticeship in the Chicago StockYards he began to study with me in 1929.Schoeffel is interesting, energetic, and loyal.He was then and still is, somewhat of a mysticand inclined in ordinary conversation to quotefreely from Goethe 's Faust, Shakespeare, andthe Bible, as well as other primary sources.In 1933 his spoken English was not onlystrongly guttural, but also very earthy, punc-tuated frequently with Schwabian German.

98



ANTICOAGULANTS: A HISTORICAL SYMPOSIUM

After farmer Carlsoii left, Schoeffel stormedback and forth in the laboratory shouting,"Vat da Hell, a farmer shtruggles nearly 200miles in dis Sau-wetter, driven by a shpectreand den has to go home vit promises datmight come true in five, ten, fifteen years,

maybe never. Who knows? 'Get some goodhay-transfuse.' Ach!! Gott, how can you dodat ven you haf no money?" he snarled.He dipped his hands into the milk can re-

peatedly and while rubbing them muttered,"Dere's no clot in dat blook! BLUT, BLUTVERFLUCHTES BLUT. 'Die Menschen dau-ern mich in ihren Jammertagen.' " (FaustProlog., line 297) and then, "Vat vill he findven he gets home? Sicker cows. And ven he andhis good voman go to church tomorrow andpray and pray and pray, vat vill dey haf on

Monday? MORE DEAD COWS!! He has no

udder hay to feed-he can't buy any. Andif he loses de bull he loses his seed. MeinGott!! Mein Gott!! Vy didn't ve anti-shi-patedis? Ya, ve should haf anti-shi-pated dis."We took the blood and hay and played about

with them until about 7: 00 p.m. when I

headed for home. As I left the laboratory,Schoeffel grabbed me by the shoulders, lookedme squarely in the face and said, "Before you

go let me tell you something. Der is a desh-tiny dat shapes our ends, it shapes our ends Itell you! I vill clean up and gif you a docu-ment on Monday morning. "

Development of the BioassayTwo fundamental issues confronted us.

First there were no chemical criteria availableto establish the presence of the hemorrhagicagent. Therefore, a bioassay involving a

small experimental animal (rabbits) ) offeredthe only practical means of appraising theanticoagulant activity of test hays and ex-

tracts prepared therefrom.It was clear from the pioneer papers of

Roderick that the sweet clover disease was

completely reversible. The eating of spoiledhay, even over long periods, caused no per-manent functional change, no demonstrablemorphologic change, and no detectable patho-logic change of the liver, the assumed primarysite of prothrombin synthesis. Nevertheless,

the immediate prospects of developing a reli-able and simple bioassay were not bright; in-deed they were dark, "dark like the inside ofa cow. " We had not had previous experiencewith that complex problem-blood coagulation.

Schoeffel and Roberts first showed that theHowell method for estimating prothrombinactivity did not have the precision required.Smith and Roberts showed that the whole bloodcoagulation time was too variable, and thatthe Quick 1-stage method using whole plasmaleft much to be desired. Smith also showedthat there was a wide variation in the responseof individual rabbits to the standard dose of50 Gm. of the spoiled hay. So Campbell andSmith bred and reared a susceptible rabbitcolony specifically for the assay.At that time, 1935-1938. a bloody and amus-

ing polemic raged among the coagulation spe-cialists on how to estimate "prothrombinconcentration or activity' '-whether it shouldbe done by the 1-stage method of Quick* orthe 2-stage method of H. P. Smith and co-workers.'-3, 8 We tried to keep out of thatbrawl. In 1938 Campbell finally got over thechief obstacles. He adapted the Quick 1-stagemethod to our conditions, primarily by relyingon the clottability of diluted plasma withinthe concentration range 12.5 to 8.34 per cent.He eliminated some of the inherent daily vari-ations by fasting the assay rabbit 24 to 36hours before feeding any preparation undertest, by making the plasma clottability testspromptly after drawing the blood, and bycomparing the test plasma against the normalplasma of each rabbit.Through the use of individually standard-

ized rabbits (the standard response being that

*The intricacies of the blood coagulation phenom-enon are outside the scope of this discourse. Sufficeit to state that it is now accepted by most "coagula-tionists'' that a prolonged Quick 1-stage ''prothrom-bin time " (when the fibrinogen is normal) inducedby Dicuniarol and the like is a primary deficiencyin factor VII and prothrombin. See British MedicalBulletin, vol. II, no. 1, Blood Coagulation and Throm-bosis, Medical Department, The British Council, Lon-don, November (1955), and the lectures by OwrenP. A. on Coagulation of Blood, etc., Northwest Medi-cine, January, pp. 31-39, February, pp. 159-166, andMarch, pp. 298-307, 1957.

!)!}



LINK

induced by the anticoagulant in 50 Gm. ofspoiled hay) and by having the assay on astrictly differential basis the ever presentproblem of biologic variation was greatly re-duced.Some side observations were made by Camp-

bell on the plasma of rabbits fed the spoiledhay or fractions thereof that were later re-ported by others. A plasma factor beyondthat needed by the classical blood coagulationexpression of Morawitz-Field and Spero washinted at in one of Campbell's reports. Butthese hares were not hunted. Our goal wasto make real a substance that abolished theclottability of cattle blood in agriculturalpractice. To use the vernacular, the bioassayusing the 1-stage plasma clottability was al-tered so that " it worked," and few of thevaluable assay rabbits were lost in the process.One of them known as Bess Campbell wasused for about 200 individual assays, over aperiod of 5 years.

Isolation, Crystallization, Identification, andSynthesis of Dicumarol

Between that fateful Saturday in February1933 and June 1939 a long and arduous trailwas followed by Smith, Roberts, and especiallyCampbell, to lay the anticoagulant out on thebench. I would like to detail some of thechemical extraction, separation, and isolationproblems that the spoiled sweet clover haypresented. This hay was indeed a kind ofbiochemical grab-bag and yielded many inac-tive products, some new, most of them old.But suffice it to state that many a seethingand simmering hope did not become reality.At times the hemorrhagic agent appeared tohover before us like thistle down only to eludeus like the will-o-the-wisp. At one time it wasthought to be a porphyrin-like substance, apheophytin resulting from the degradation ofthe chlorophyll in the spoiling process.

Finally in the dimness of dawn on June 28,1939, after working all night, Campbell sawon a microscope slide what turned out to becrystalline Dicumarol. Two hours later hehad collected about 6.0 mg. of it.

When I reached the laboratory that morn-ing Campbell was asleep on the laboratorycouch; the door to the room was guarded byone Chet Boyles, a soldier of fo-tune on theW.P.A. relief roles, who assisted Campbellwith the bioassays. Boyles was an excellenthandler of animals for he had served 2 yearsas helper to a veterinarian before he came tous.As I walked into the room, Boyles was tak-

ing a nip from the contents of a bottle whosebottom layer consisted of carpet tacks, theupper layer of 95 per cent ethanol. Withoutthe flicker of an eyelash Boyles said to me,"I'm celebrating, Doe. Campy has hit thejack-pot." (As though I didn't know that hehad been hitting that bottle for months.)But Boyles' surmise was correct this time.

Campbell did have Dicumarol and the firstbioassay to establish its anticoagulant potencywas already in process!

Campbell avoided me for 2 days-until theresults of the assay were available-and thenhe came in to report.

There is a bed-rock of matter-of-fact com-mon sense in Campbell 's makeup. He was notinclined to show his emotions, but it was ap-parent that he was secretly as happy as a boywho had just caught his first big fish. Hepassed the vial to me and said, " This isH. A.!" (H. A. was the laboratory code forhemorrhagic agent.) I did not disclose thatBoyles had given me the tip-off. I told Camp-bell that I knew a couple of lines of Germanpoetry that fitted the occasion, and I recitedto him,

"So halt'ich's endlich denn in meinemHainden

Und nenn' es in gwissen Sinne mein."

We sent a short wire to Schoeffel, who wasthen in the control laboratory of the AmericanMedical Association in Chicago. He respondedat once with a 200-word reply wherein he ex-pressed his complete confidence in Nature,Fate, and us.Mass isolation was started at once, and a

stock of about 1,800 mg. of the crystallineanticoagulant was accumulated (Stahmann).

1.00




The problem of determining its structure fellto the sensitive, brilliant, and deft C. F. Hueb-ner, who with some assistance from his livelyimagination made the correct structural diag-nosis as 3, 3'-methylenebis (4-hydroxycouma-rin). He set the sights for the synthesis,which was achieved on April Fool's day, 1940.The synthetic and the natural product wereshown to be chemically identical. Subse-quently, Overman and Sullivan, throughcarefully conducted tests on the rabbit, rat,guinea pig, mouse, and dog, hall-marked thenatural and synthetic products as biologicalequals.The determination of the structure of the

anticoagulant as a 3-substituted derivative of4-hydroxycoumarin makes it appear that bothof the undesirable aspects of the commonsweet clovers-their unpalatability (bitter-ness) in the green state and the tendency ofthe hays to cause hemorrhage when improperlycured-have a common basis in the coumarinmolecule. The biological synthesis duringspoilage can be rationalized as an oxidation ofcoumarin to 4-hydroxycoumarin which uponcoupling with formaldehyde leads to Dicuma-rol.

I-0coumarin

OH H HO

+

CH20* CXH m

4-hydroxy- 3,3-methylene bis(4-hydroxy-coumorin coumorin)

Physiologic Action of DicumarolAfter synthetic Dicumarol became available

in quantity the essentials of its physiologicaction were quickly established. It was shownthat there is a lag in response, a variation inthe intensity and duration of the hypopro-thrombinemia (plasma prothrombin clottingtime), depending on the size of the dose. Ineach species tested a certain single dose levelgives the most efficient response. Below thislevel the efficiency of action is decreased by a

threshold effect and at high levels by incom-plete absorption of the drug.Due to the latent or lag period of 12 to 24

hours before the drug 's action becomes ap-

parent, there is a cumulative effect followingrepeated administration. Thus it was antici-pated that in clinical practice this action will

vary with the individual and because of thisvariation optimal therapeutic effects withouthemorrhage would be obtained only when thedosage is individualized.A brief summary1-3 of the details follows:1. There is a wide species difference in the

response induced in the rabbit, rat, guineapig, mouse, dog, cat, and chicken, and thisvaries with the age and sensitivity of each in-dividual. Broadly speaking, the rat and mouseare the most sensitive, the cat and dog inter-mediate, and the rabbit, the cow, and thechicken the least sensitive.

2. The vitamin K and C levels in the dietaffect not only the intensity but also the dura-tion of the anticoagulant action. I proposeto elaborate on this later.

3. The nutritional status of the animal af-fects the anticoagulant response-fasting gen-erally enhances it in all species.

4. Any pre-existing hypoprothrombinemialike that inducible by the salicylates (as-pirin), the sulfa drugs, or mild chloroformanesthesia augmented the response.

5. The hepatic and renal function influencesboth the intensity and duration of the re-sponse.

6. The presence of drugs that affect thetotal functioning capacity of the liver, likethe methylxanthines (theophyllin) and thedigitalis drugs, have a mild but definitely de-tectable counter action.

7. Pregnant or lactating females show aslight resistance to the drug's anticoagulantaction.

These observations did not exhaust theconditions that can influence Dicumarol's ac-tion but they cover the essential points.Finally, it should be added that in Dr. Best'sdepartment at Toronto, Dale and Jaques first9and later Meyer and co-workers10 at Wiscon-sin General Hospital, and others"-13 were ableto show that a primary relationship existsbetween thrombus formation and the clottingmechanism of the blood. These studies estab-lished for the first time that an effective re-duction of extravascular and intravascularthrombus formation parallels the diminishedhypocoagulability induced by Dieumarol. It

1.01



1LINK

was also shown by Spooner and Meyer14 that,when Dicumarol is given to dogs in safelyusable therapeutic doses, it definitely decreasesplatelet adhesiveness; at the same time Quickshowed that it also reduced platelet aggluti-nability.15 Thus the clinical use of theanticoagulant as a prophylactic agent for(against) thrombosis rested on a sound experi-mental basis.

Breaking the Bonds of the Usual Pattern ofThoughtWhen we turned Dicumarol over to the

clinicians in the years 1940 to 1942, one sig-nificant point, clearly established by our work,was at first missed, in fact denied.6 I havereference to the capital fact that vitamin K(all forms-some better than others) cancounteract the action of Dicumarol.* I em-phasized this in letters, personal conversations,and in my first lecture on Dicumarol at theMayo Clinic and at Wisconsin General Hos-pital. In spite of these efforts the first clinicalreports carried the statement that "vitaminK has no effect as an antidote to the adminis-tration of Dicumarol. " The editorial andannotation writers for the medical journals,those who only " think " but " don 't try,"innocently reiterated this statement." 2 Whilein error, the clinicians were in good company,for an authority on blood coagulation16 hadwritten in 1937, and again in his book pub-lished in 1942, that "vitamin K will not re-store the prothrombin concentration" de-pleted by Dicumarol.17

Originally these denials made me very un-happy. The misfortune of being accused oferror was not the primary basis for the unhap-

*For an account of how, in January 1939, a bulldesperately sick from eating spoiled sweet clover hay(he was "down," blood was oozing from the nose,and a massive hematoma adorned the right thigh)was rescued from the clutches of death via a vitaminK1 concentrate prepared from alfalfa hay, see refer-ence 2. Originally not even this "bull story" couldbreak the bonds of the usual pattern of thought. The

equation isSpoiled sweet clover hay

Normal bull 4 Bleeding bullVitamin K1 concentrate

a completely reversible reaction.

piness, for we were certain that the antidotalcapacity of vitamin K would in time be sus-tained in the clinic. What did disturb mewas the needless induction of the hemorrhagic''sweet clover disease" in man and the stigmatemporarily attached to Dicumarol, that itwas a dangerous drug.'8 And this did happen.A feature of science that has always ap-

pealed to me is that sooner or later, and usu-ally sooner, "the truth will conquer. "

Dr. Shepard Shapiro in New York City wasthe first clinician (February 1942) to sustainour claims that vitamin K can counteract theanticoagulant action of Dicumarol in manwhen liver function is adequate.' 20 Subse-quently he was independently supported byTownsend and Mills in Canada,21 Lehman inSweden,22 and finally by Cromer and Barker23at the Mayo Clinic, as well as others. Todayit is accepted that the water-soluble forms ofvitamin K or vitamin K, given orally cansuccessfully antidote overdosing with Dicu-marol, provided they be employed when re-versal is still possible.

Let us briefly examine why the error arose.The clinicians did not use a 1-stage prothrom-bin assay as sensitive as the one Campbelldeveloped for our experimental animals. Theywere originally conditioned to the low levelsof vitamin K effective in obstructive jaundice,biliary fistula, cholemic bleeding, etc. It wasalso thought that the menadione form of vita-min K might be toxic. Over 10 years wererequired to wipe out this error from clinicalpractice.To summarize, surmise, faulty thinking,

and not enough trying kept vitamin K frombeing the corner building stone in Dicumaroltherapy that it deserved to be from the outset.

In 1950 Marple and Wright (pages 149 and181)6 wrote, "When bleeding occurred fromthe clinical use of Dicumarol the fault restedwith the physician who administered thedrug. "

Enwthusiasmn-Muddle-ConsolidationWithin 2 years after Dicumarol was syn-

thesized, over 100 related 3-substituted 4-hydroxycoumarins were prepared in my lab-

102




oratory. Synthesis ran substantially aheadof biochemical appraisal. Accordingly, whenI gave the Harvey Society lecture on "TheAnticoagulant from Spoiled Sweet CloverHay" in January 19441 it was indicated that"it would not be valid to conclude from therelative appraisals on activity made with therabbit-that Dicumarol is the most desirablecompound for clinical use. " It was indicatedthat " In the course of the routine appraisal ofthe many compounds tested it was learned thatsome of them exhibited a slower but moresustained hypoprothrombinemic action, whilethe action of others is of shorter duration. Itwill take some time before final judgmentcan be passed on this subject. From the ex-perience gained with other pharmacologicalagents it is abundantly clear that the finaltest is the action in man under a variety ofconditions. The unpredictable can be sur-prising) so, as we see it, we might now be atthe beginning of things and not at the endin this field of study. "Being an agriculturist I have little confi-

dence in predictions, including my own. Thesituation can now be appraised in the lightof wisdom after the event. Bear in mindthat the statement quoted was made less than4 years after Dicumarol became known to usand before extensive clinical information onthe response in man was available. About50 reports on the clinical use of Dicumarol hadappeared between 1941 to 1944.5 6*The appearance of any new drug creates ani

interesting cycle of events, and Dicumarolwent through that cycle quite rapidly. Thefirst preliminary reports indicated that an at-mosphere of optimism prevailed. They evoked

*The first clinical report to appear was by Butt,H. R., Allen, E. V., and Bollman, J. L.: Preparationfrom spoiled sweet clover (3,3'-methylene-bis-4-hy-droxycoumarin) which prolongs the coagulation andprothrombin time of blood: Preliminary report ofexperimental and clinical studies, Proc. Staff Meet.Mayo Clinic 16: 388-395 (June 18), 1941. See alsoAllen, E. V., Barker, N. W., and Waugh, J. M.,J.A.M.A. 120: 1009, 1942; Wright, I. S., and Pran-doni, A., J.A.M.A. 120: 1015, 1942; Bingham, J. B.,Meyer, 0. 0., and Pohle, F. J., Am. J. M. Sc. 202:563, 1941.

prompt favorable editorial comment in theLancet (September 13, 1941) under the title,"Heparin and a Rival." Then came the see-ond period-a period of muddle. Enthusi-asts and skepties for anticoagulant therapywith Dicumarol were created, and it can bestated that some of the skepties condemnedthe drug in no uncertain terms, though theywere largely armed with surmise, faulty, orno prothrombin clotting time determinationsand they used the antidote vitamin K inade-quately. Then came the third period of con-solidation, from which it can now be con-cluded that a better anticoagulant of the Di-eumarol type was desired.

Since Dr. Wright asked for aspects of hu-man interest, let me add another slice fromthe Dicumarol story. Early in September1945 I was fed up with laboratory work, etc.,and I went off on a canoe trip with my family.On this trip we were caught in a cold railistorm. I got soaked and overexhausted. Twoweeks later I came down with what I had hadonce before-after a similar heavy physicalbout, as a student in Switzerland-wet pleu-risy. At first my doctor thought I had pneu-monia; then I told him about my previousbouts of tuberculosis; so the diagnosis waschanged to reactivated pulmonary tuberculo-sis. I spent 2 months at Wisconsin GeneralHospital and then was transferred to Lake-view Sanatorium headed by the double crossof Lorraine. Here I was supposed to vegetatelike a topped carrot. I did rest there, physi-cally for 6 months, took nothing stronger thancod liver oil and 3 bottles of beer a day, butkept the aged tuberculosis out of mny mindby studying laboratory records and readingthe history of rodent control from ancient tomodern timeS.24, 25

A "Janus" in the Coumarin FamilyNow brace yourselves, for I propose to shift

from a "cow poison" that had become a drugof substantial clinical usefulness, to a "ratpoison" converted to a drug, which has I be-lieve most of the desirable features that canbe expected from an anticoagulant to be givenprimarily via the oral route.

103



1LINK

The many coumarins synthesized between1940 and 1944 were listed by numbers in logi-cal groups based on their chemical structure.'While I was in the sanatorium in 1945-1946the laboratory work was practically at a stand-still. There were few students available, sincemost of them were still in the armed forces.So I had ample time to reexamine all thechemical and bioassay data available. Uponthe return of L. D. Scheel from service in thespring of 1946, he was assigned to the task ofreappraising the anticoagulant activity of thecompounds numbered from 40 to 65. Theywere made by Ikawa in 1942-43. Instead ofusing only rabbits for the bioassays Scheelalso used rats, mice, and dogs. In 1946-1948he defined coumarin numbers 42 and 63 as

being much more potent than Dicumarol inthe rat and dog, as capable of producing a

more uniform anticoagulant response, and as

having the quality of maintaining a more se-

vere state of hypothrombinemia without in-ducing visible bleeding. Certain chemicalproperties were also considered: the degree ofpurity readily attainable (absence of taste andodor), the cost of making them, and the prop-

erty of being convertible to stable water-sol-uble salts.Back in 1940 to 1942, Overman, Field, and

my colleague, C. A. Bauman, had studied ex-

tensively the action of Dicumarol in the lab-oratory rat, and the effect of diet on theresponse, specifically the influence of vitaminK and foods rich in it. Later in 1942 I per-

sonally, with the help of good old Schoeffel,set up field trials to ascertain the suitabilityof Dicumarol for rodenticidal purpose. It was

concluded that the activity of Dicumarol inlthe rat was not high enough to make it prac-

tical for rodent control. This was found tobe largely due to the vitamin K content ofmature grains and the availability of green

foods with a high vitamin K content. It was

shown that rats could tolerate a daily intakeof 2.0 mg. of Dicumarol for 60 or more daysdue to the vitamin K content of the naturalfoods available. On a semisynthetic diet es-

sentially free from vitamin K the survival

time was about 15 to 23 days. When 5 mg. of

vitamin K per day were added to the artificial

diet, the rats also tolerated 2.0 mg. of Dicu-marol daily for over 60 days.

Early in 1948 I told Scheel and DorothyXXu that I wanted to propose no. 42 forrodenticidal use.24' 25 This proposal shook thelaboratory. I can sum up by stating the con-sensus of opinion "the boss has really goneoff the deep end this time." Scheel favoredno. 63 for clinical purposes. They are chem-ically closely related, no. 63 being a directderivative of no. 42. To make a long storyvery short, early in 1948 no. 42 was promotedfor rodent control under the auspices of theWisconsin Alumni Research Foundationthrough the able, enthusiastic, and public-spirited Ward Ross, General Manager of thisorganization. Within a short time this effortrevolutionized the art of rodent control (mul-tiple doses as opposed to the single dose ofthe highly toxic poisons), and warfarinrapidly became and still is the leader in therodenticide field.* The name Warfarin wascoined by me by combining the first letters ofthe Wisconsin Alumni Research Foundationwith the " arin " from coumarin-and it is nowa household word throughout the world.**Between 1948 and 1952 Dicumarol was, so

to speak, being squeezed by chemical kin stem-ming primarily from European studies.26"Imitation is the sincerest form of flattery."'Curiously, one of them, a derivative of Dicu-marol, trade-named Tromexan, was not seri-ously considered by us as early as 1940.Though it acted somewhat faster than Dicu-marol, it required substantially larger doses

*Just as cattle eat hemorrhagic sweet clover hayuntil they die without visible sensory responses, therat eats warfarinized cereal grain bait until fatalhemorrhage sets in. Neither bait refusal nor baitshyness develops. Indeed, the rodent 's departure isbiblical "death without sting." In Maxwell Ander-soi 's drama, Elizabeth the Queen says, "'To theend of time it will be so . . . the rats inherit the

earth.'" Since warfarin has become available, thisneed not be so. Furthermore, via the water-solublewarfarin sodium the rat can drink unto death.

**Warfarin is the safest rodenticide known. Upto now, in the United States there is no recorded case

of a warfarin-induced fatality in man, although over

140,000,000 pounds of warfarin containing bait (0.025per cent) have been distributed since 1950.

1.04




to get the e(uivalenit anticoagulant action.*The second, Marcumar, a close kin to warfariii,was also passed by us, since its water-solublesodium salt is less stable than warfarin so-dium. Milligramn for milligram, Marcumar ismore active than warfarin and its action isalso more prolonged. But as a result of theclaimns made about Tromexan and Mareumar

*The clinical promotion of Tromnexan (bis-3,3'-4-oxycoumariniyl) ethyl acetate, referred to as B.O.E.A.in the article by Burt, C. C., Wright, H. P., andItubik, M., Brit. M. J. 2: 1250, 1949, precipitatedinteresting editorial comment under the heading,J)angers of Diciemarol (pp. 1279-1280). In this edi-torial it was suggested that since Tromexan seemedto be superior to Dicumarol "'owing to its shorter-lived action . . . '' and in view of recent reports ofthe drug 's (Dicumnarol 's) efficiency as a rat poison,it may be that Dicumnarol will ultimately be moreuseful for that purpose.'"

Unfortunately the significance of our paper oln thelaction of Dicuniarol in the rat dealing specificallyw ith the effect of diet and vitamin K on the anti-coagulanit action (J. Nutrition 23: 589-602, 1942) wasnot appreciated by O'Conner, J. A., Research 1: 334,1948, who suggested the use of Dicumarol for rodentcontrol. Had O'Conner read our paper carefully, hewould not have made this suggestion. The criticalissue is that Dicumnarol's anticoagulant action in therat subsisting on natural grain foods is too slow tobe practical. The level of Dicuinarol in the bait hasto be set so high that other ammimnals (cat, dog) andtilmildreni (accidental ingestion) would be vulnerable.

It was the inefficiency anid slow-ness of Dicumarolto kill rats under practical field conditions that causedme not to suggest its use as a rodenticide in 1941-1943 (letter, Link, K. P., to the National DefenseResearch Council, Washiingtonm, D.C., dated March10, 1943, and confidential disclosures, 1942-1943,to the late Professor Homer Adkins and ProfessorH. Gilman, official investigators and project leadersof N.D.R.C. and O.S.R.D. (confirmatory letterof Gilinan to Link, June 11, 1952). Instea(d of Dicu-mnarol the amuck more potent and efficient (no. 42)warfarin was recommended. Nevertheless, O'Conner Ispaper served a useful purpose in rodenticide controlcircles, and he must be accredited with being the firstone to stimulate, via the printed page, the backwardpest control workers by pointing out the potentialsof anticoagulants (Link, K. P., letter December 6,1948, to U.S.D.I. Fish and Wildlife Service, Denver,Colo.). I had attempted to create an interest inwarfarin via letters and memoranda, which at firstfailed to reach the objective (see reference 24 andparticularly reference 25). A complete history of thewarfarin development based on 10 years of practicalfield experience is in the process of being prepared.

I took another good look at the niass data inthe light of what clinicians were seeking,namely an anticoagulant that could be usedvia any route with the retention of the virtuesof Dicumarol but without its limitationss.

OH HO

t JCH2- t

Dicuma rote

CH3C.O

ONo CH2

X CH-¢

Warfarin Sodium(Coumod.n SodiunAl

2

QHC-OEt HOr ICr~~r-H- N

Tromexone

CH3OH CH2

McH-mo

M arcurmaote

Based in part on the supposition that theresponse of the rat to warfarin (no. 42) was avery reliable index of how man would respond,late in 1950 I told Dr. S. Shapiro and Dr. 0. 0.Meyer that the water-soluble sodium salt ofwarfarin should be tried oil ilan. In 1941the clinicians had literally snatched the "cowp0oiso01 from us, but the transition to a sub-staiice originally promoted to exterminate ratsaimd mice was a bit more than they could ae-cept with real enthusiasm. Then, oil April 5,1951, we were informed by Captain J. Love(MC) in the U.S.N. at Philadelphia that anarmy inductee was admitted to the NavalHospital who had taken over a period of 5days a concentrate of warfarin designed forrodent control.27 The package contained 567mng. of warfarin in corn starch. The inducteehad followed the multiple dose directions onthe package. It became clear to him thatwarfarin was not anl efficienmt agent "to shuffleof " this " mortal coil. " It allowed too muchtime for thinking-so he went to the hospitalwith a fully developed case of hemorrhagic"ssweet clover disease." He was treated perthe direetions-blood transfusion and largedoses of Vitamin K-and made an uneventfulrecovery. 28

This incident acted as a catalyst. Shapiro02and Mleyer3 both concluded from their care-fully done work with warfarin sodium thatit did possess certain properties not inherentin Dicumarol or the other anticoagulants they

105



LINK

had tried. After Collin Schroeder perfectedthe process of making warfarin sodium, Iinduced my long-standing friend, Dr. S. M.Gordon, of the Endo Laboratories, RichmondHill, N.Y., to make it available for clinicaluse. This he did, under the trade nameCoumadin Sodium. Today it would appear,from the 15 to 20 clinical papers on warfarinsodium that have been published (see refer-ence 32), that most of the drawbacks ofDicumarol have been overcome. Warfarinsodium is at least 5 and possibly 10 timesmore potent than Dicumarol. It is the onlysynthetic anticoagulant available today fortherapeutic anticoagulation that can be givenorally, intravenously, intramuscularly, or rec-tally.32 The rate of absorption is almost thesame, irrespective of how it is administered.No other anticoagulant of the Dicumarol typehas all these virtues. Of course, an overdosagecan be readily corrected via vitamin K. Itacts faster than Dicumarol, and fewer pro-thrombin times are required in its routineuse. To use the words of both Shapiro andMeyer, "It is easier to handle clinically. " Itis my firm belief that in time it will replaceDicumarol on the basis of its performanceover a wide variety of conditions and thatother anticoagulants of the Dicumarol typewill not be superior.

It always seems appropriate to me to vis-ualize successful anticoagulant therapy withthe Dicumarol-type drugs as being shaped likea triangle with accurate "prothrombin as-says " at one corner, vitamin K at another,and sound clinical judgment at the third.

Vitamin K(water-soluble and KJ)

Reliable clottingtime assays

Clinical judgment

Each corner is linked to the other by wayof the connecting sides. There should be noseparation, each is vitally dependent on theother two. Though the clinical judgment begood and the "prothrombin time" accurate,the vitamin K corner might still have to beevoked, since each individual patient is essen-tially "an unstandardized biologic entity,"errors in dosage can be made by the hospitalservice, the patient might have a silent ulcer,or the functioning of the liver or kidney mightunknowingly be penumbral.On September 29, 1955, I got a card from

a former Wisconsinite working in Fitzsi-mons Army Hospital in Denver, Coloradowhich read, "The President is getting oneof your drugs and it's not Dicumarol." Aday later press secretary J. C. Hagerty an-nounced,33 "The heparin which was used ini-tially as the anticoagulant has been replacedby a drug of the Dicumarol type. The presentprothrombin level has been well maintained. "I knew of Colonel Pollock's paper, "Clinicalexperience with warfarin (coumadin sodiuma new anticoagulant)," read before the firstannual meeting of the American College ofAngiology Atlantic City, N.J., on June 4,1955, and I surmised that the most importantman in the world today was being anticoagu-lated via warfarin sodium.34 This surmiseproved to be correct and since then it is anopen secret that warfarin sodium was beingused. "The unpredictable can be surprising."

In closing I wish to indicate that what mylaboratory has achieved in the past 21/2 dec-ades represents the combined effort of manystudents. It is fun to be the reporter ornarrator of this highly successful adventure.To use the words of the late Allan Gregg,35my students represented much "emergentability." I think the secret of their successis 3-pronged: they never ceased to wonder,they kept on trying, and they were on aproject directed toward doing mankind somegood instead of trying to destroy it.

REFERENCES1. LINK, K. P.: The anticoagulant from spoiled

sweet clover hay. Harvey Lectures Series34: 162, 1943-44.

106




2. -: The anticoagulant Dicunmarol. Proc. Inst.Med. Chicago 15: 370, 1945.

3. -: The anticoagulant 3,3'-methylene-bis (4-hydroxycoumarin). Fed. Proc. 4: 176, 1945.

4. ANON.: New York Heart Association Quar-terly, New York, vol. 2, 6-11, 1950.

5. SHAPIRO, S., AND WEINER, M.: Coagulation,Thrombosis and Dicumarol. New York,Brooklyn Medical Press, 1949.

6. MARPLE, C. D., AND WRIGHT, I. S.: Throim-boembolic Conditions and Their Treatmentwith Anticoagulants. Springfield, Ill.,Charles C Thomas, 1950.

7. WRIGHT, I. S., MARPLE, C. D., AND BECK,D. F.: Myocardial infarction (Report ofCommittee on Anticoagulants). Publishedfor the American Heart Association. NewYork, Grune & Stratton, 1954.

8. LINK, K. P.: Dicumarol and the Estimationof Prothrombin. Transactions of the FirstConference on Blood Clotting and AlliedProblems. New York, Josiah Macy, Jr.Foundation, 1948, p. 128.

9. DALE, D. U., AND JAQUES, L. B.: Preventionof experimental thrombosis by dicoumarin.Canad. M.A.J. 46: 546, 1942.

10. THILL, C. J., STAFFORD, W. T., SPOONER, M.,AND MEYER, 0. 0.: Hemorrhagic agent 3,3'-methylene-bis (4-hydroxyeoumarin); its ef-fect in prevention of experimental throm-bosis. Proc. Soc. Exper. Biol. & Med. 54:333, 1943.

11. RICHARDS, R. K., AND CORTELL, R.: Proc. Soc.Exper. Biol. & Med. 50: 237, 1942.

12. -, AND -: Proc. Soc. Exper. Biol. & Med.52: 358, 1943.

13. BOLLMAN, J. L., AND PRESTON, F. W.: Effectsof experimental administration of dicou-marin 3,3'-methylene-bis (4-hydroxycoumar-in). J.A.M.A. 120: 1021, 1942.

14. SPOONER, M., AND MEYER, 0. 0.: Effect ofDicumarol [3,3'-methylene-bis- (4-hydroxy-coumnarin) ] on platelet adhesiveness. Am.J. Physiol. 142: 279, 1944.

1.5. BORONOFSKY, I. D., AND QUICK, A. J.: Heparinand agglutination of platelets in vitro.Proc. Soc. Exper. Biol. & Med. 53: 173,1943.

16. QUICK, A. J.: Coagulation defect in sweetclover disease and in hemorrhagic chick dis-ease of dietary origin; consideration ofsource of prothrombin. Am. J. Physiol. 118:260, 1937.

17. -: The Hemorrhagic Disease and the Physi-ology of Hemostasis. Springfield, Ill.,Charles C Thomas, 1942.

1.8. BIGGS, R., AND MOFARLANE, R. J.: Estima-tion of prothrombin in dicoumarin therapy.Brit. M. J. 296, 1949.

19. -, AND -: Human Blood Coagulation andIts Disorders. Oxford, Blackwell, 1953.

20. SHAPIRO, S., REDISH, M. H., AND CAMPBELL,H. A.: Proe. Soc. Exper. Biol. & Med. 52:12, 1943.

21. TOWNSEND, S. R., AND MILLS, E. S.: Effectof synthetic haemorrhagic agent 3,3'-methy-lene-bis (4-hydroxycoumarin) in prolong-ing coagulation and prothrombin time inhuman subject. Canad. M.A.J. 46: 214,1942.

22. LEHMANN, J.: Thrombosis: Treatment andprevention with methylene-bis- (hydroxy-coumarin). Lancet 1: 611, 1943.

23. CROMER, H. E.. JR.. AND BARKER., N. W.: Ef-fect of large doses of menadione bisulfite(synthetic vitamin K) on excessive hypo-prothrombinemia induced by dicumarol.Proc. Staff Meet. Mayo Clin. 19: 217, 1944.

24. LINK, K. P.: The Rat. Paper read before theMadison Literary Club, Madison, Wis.,January 9, 1950. Filed in the WisconsinHistorical Library, Madison, Wis.

25. DE KRULF, P.: Sure Death to Rats. In Reader'sDigest. Pleasantville, New York, Reader'sDigest Association, March 1951.

26. KOLLER, T., AND MERZ, W. R., Editors: I.International Conference on Thrombosisand Embolism. Basel, Switzerland, B.Schwabe, 1954.

27. HOLMES, R. W., AND LOVE, J.: J.A.M.A. 148:935, 1952.

28. KELLUM, J. M.: Warfarin for suicide.J.A.M.A. 148: 1443, 1952.

29. SHAPIRO, S.: Warfarin sodium derivative(coumiadin sodium) : Intravenous hypopro-thrombinemia-inducing agent. Angiology 4:380, 1953.

30. -: Hypoprothrombinemiia-inducing activityof Warfarin sodium (coumadin sodium). J.Kansas M. Soc. 55: 687, 1954.

31. CLATANOFF, D. V., TRIGGS, P. 0., AND MEYER,0. 0.: Clinical experience with coumarinanticoagulants Warfarin and Warfarin so-dium. Arch. Inst. Med. 94: 213, 1954.

32. SHAPIRO, S., AND CIFERRI, F. E.: Intramnus-cular administration of the anticoagulantWarfarin (eoumadin) sodium. J.A.M.A.165: 1377, 1957.

33. HAGERTY, J. C.: Associated Press release aspublished in The Capital Times, Madison,Wis., September 30, 1955.

34. POLLOCK, B. E.: Clinical experience withWarfarin (counmadin) sodiuni, a new anti-coagulant. J.A.M.A. 159: 1094, 1955.

35. GREGG, A.: Emergent ability. In The Pharosof Alpha Omega Alpha, vol. 15, p. 3, 1951.Published at Slaterville Springs, New York.

1()7



Documents

The Discovery of Dicumarol andwilliams.medicine.wisc.edu/warfarin_discovery.pdf · 2011-11-25 · 98INKL headed the Biochemistry Department of the University of Minnesota. Hehad offered