POSTGRAD. MED. J. (1965), 41, 663.

ATRIAL FIBRILLATION IN THYROTOXICOSISTREATED WITH RADIOIODINE

J. S. STAFFURTH, M.D., M.R.C.P.,Consultant Physician

M. C. GIBBERD, M.B. (N.Z.), M.R.C.P.,Clinical Assistant, Radioisotope Laboratory

P. J. HILTON, M.A., M.B., B.Ch.,lately House Physician

Lewisham Hospital, London, S.E.13.

RADIOACTIVE Iodine (131I) is frequently recom-mended as the treatment of choice for thyro-toxicosis in patients over 45 years old withoutsufficient appreciation of the difficulties inmanagement that occasionally occur. The in-creasing frequency of post-radiation myx-oedema is now widely recognised (Green andWilson, 1964) but less emphasis has been paidto the situation in some subjects in whom thefirst therapeutic dose may have little effect. Alsolittle attention has been given to the manage-ment of a atrial filbrillation (A.F.) in thyro-toxicosis treated with 1311, though most of thesecases are now treated by this method. Thisis in contrast to the interest shown by thesurgical pioneers including Dunhill (1930) whoconsidered that reversion of A.F. to normalrhythm was a good clinical index of the effect-iveness of thyroidectomy.

In this paper patients with A.F. and thyro-toxicosis who had been referred for 1311 treat-ment are reviewed with particular referenceto reversion to sinus rhythm, the ultimatecardiac status and the risk of embolism.

Material and MethodsIn the six years from July 1958 to June 1964, 170

patients with thyrotoxicosis have been treated withradioactive iodine at Lewisham Hospital. The dosagehas been determined on the lines recommended byMacGregor (1957) whereby patients with small orimpalpalble glands have been,given 4 to 6 mc., thosewith moderate goitres 8 to 10 mc., those with largegoitres from 12 to 15 mc; in two very severe cases,25 mc. were given in dividend doses initially.Eighteen patients had been made euthyroid with

carbimazole before ""lI treatment after which thedrug was continued in decreasing dosage for threemonths. In the last two years 13 selected patientswere given carbimazole for short periods startingone week after the "'1I to ensure that severe dis-ability did not continue unnecessarily.A follow-up has also been undertaken on 19

patients with A.F. out of a total of 86 with thyro-toxicosis treated with l31I at St. Thomas's Hospitalby one of us in 1956-57 with particular reference tothe possibility of the subsequent occurrence of acerebrovascular accident.

ResultsOf the 170 subjects in the main series 39

(23%) had A.F. confirmed by electrocardio-gram (EOG) at some stage of their illness. Ineight subjects the A.F. was paroxysmal beforedefinitive treatment was given and in one itoccurred for the first time three months afterthe first dose of 131j. The age distribution ofthese cases and the number with A.F. isdepicted in Fig. 1, from which it is evidentthat A.F. is relatively uncommon under 55years but thereafter there is a marked increasewith advancing age. The average age of thesubjects with A.F. was 62.5 years, range 37to 77 years. 28 subjects were male but therewas no apparent difference in the presentation,incidence of A.F., severity or response to treat-ment between the two sexes. 19 subjects hada toxic nodular goitre of whom eight hadA.F. and one 2:1 heart block.

Reversion to Sinus RhythmThe age distrilbution of the patients with

established and paroxysmal A.F. and thenumber that reverted to sinus rhythm isdepicted in Fig. 2. Of the 39 patients withA.F. one, with a severe hereditary cerebellardegeneration, died two days after the administ-ration of 131J; 19 had persistent A.F. and 19had sinus rhythm after the completion of treat-ment. The possible role of various clinicalfeatures, including the size of the goitre, lengthof history of A.F., presence of cardiac failureand associated heart disease, and response totreatment are summarised in Table 1.

Seven of the eight sulbjects wi'th paroxysmalA.F. were ultimately in sinus rhythm, but theother, who developed persistent A.F. is brieflydescribed:Case No. 1. A female aged 64 was referred from

another hospital with a 2-year history of thyro-toxicosis and the recent onset of paroxysmal A.F.She was given 8 mc. "1I and in three months thethyrotoxicosis was improved though not cured, butA.F. had now become estalblished. A further 4 mc.

by copyright. on June 12, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.41.481.663 on 1 Novem

ber 1965. Dow

nloaded from

664 POSTGRADUATE MEDICAL JOURNAL November, 19!

60

*aA.F

zw

30-

w

z 20

10-

<45 -55 -65 -75 >75E (years)

FIG. 1. Age distribution of 170 patients withthyrotoxicosis treated with radioiodine, and thenumbers with atrial fibrillation.

"'lI were given after which she soon became euthy-roid but A.F. persisted and was still present tw oyears later.

Eight subjects with A.F. had been treatedwith carbimazole before 3'II was given. Ofthese four had reverted to sinus rhythnm beforethe 1:3J while the other four, one with severerheumatic heart disease, all have persistent A.F.Three subjects with A.F. were given carbima-zole for the first tinme after the 1l, all ofwhonm reverted to sinus rhythm. One patientwho reverted to sinus rhythml on carbimazoledeserves brief nmention, as A.F. developed again'hree months after the initial dose of "'11 whenthe carbimazole was stopped.Case No. 2. A female aged 62 was admitted to

another hospital with cardiac failure and A.F. dueto thyrotoxicosis. The cardiac failure responded tothe usual measures and she wvas referred for "3'Itreatment. As she was a severe case it was decidedto treat her initially wvith carbimazole and in six

weeks she reverted to sinus rhythm. Four monthslater when she was euthyroid, 11 mc. 13' I ere givenfour days after the carbimazole had been stopped.This was recommenced nine days later and con-tinued in gradually reduced dosage for three monthsduring which she remained euthyroid. Symptoms ofthyrotoxicosis then soon returned with the reoccur-rence of A.F. Carbimazole was again given withearly return to sinus rhythm and after a further6 mc. 13'I she became euthyroid and remained insinus rhythm wvhen the carbimazole was finallystopped.

Onset of Atrial Fibrillation after I'll Administ-rationThree subjects are known to have developed

A.F. again, after the initial dose of 1311 hadbeen followed by a period of sinus rhythm.One other developed A.F. for the firs,t timeduring a period of emotional strain, threemonths after the first dose of 1311. Case 2 hasbeen described already but two others deserve

by copyright. on June 12, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.41.481.663 on 1 Novem

ber 1965. Dow

nloaded from

November, 1965 STAFFURTH, GIBBERD and HILTON: Atrial Fibrillation

E - estcblished A.F before treatment

p = paroxysmal A.F 9"

12

zw

,L 6-0~

Dz 3-

665

D2 - persistent A.F dter treatmnt

* 5 sinus rhythm

E

IP7

1%

<45 s55 -65 -75 > 75GE (years)

FIG. 2.-Age distribution of thyrotoxic patients withestablished or paroxysmal atrial fibrillation,together with the cardiac rhythm after completionof treatment. N.'B. * patient who died two daysafter the administration of 131I.

TABLE 1SOME CLINICAL FEATURES OF PATIENTS WITH ATRIAL FIBRILLATION INRELATION T HEART RHYTHM AFTER COMPLETION OF TREATMENT WITH 1311.

N.B. One patient who died shortly after the administration of 1311 has notbeen included.

Clinical Features final Rhythm Total~~~~~~Sinus__A.F. __ _ _ _ _ _ _ _ _ _

Moderate or large goitre 5 9 14Impalpable or mall goitre 14 10 24Toxic nodular goitre 2 6 8Paroxysmal A.F. 7 1 8A. F. present > 6 months 4 10 14A. F. present < 6 months 9 5 14Cardiac failure on presentation 10 8 18Associated heart disease present 4 2 6Euthyroid with one dose of 1311 8 9 17Euthyroid with > one dose o£f isIl 6 10 16Hypothyroid after 131I 5 5

O -

by copyright. on June 12, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.41.481.663 on 1 Novem

ber 1965. Dow

nloaded from

666 POSTGRADUATE MEDICAL JOURNAL November, 1965

TABLE 2

CARDIAC STATUS OF PATIENTS WITH ATRIAL FIBRILLATION AFTER COMPLETIONOF TREATMENT WITH 1311.

N.B. One patient who died shortly after the administration of 131I has notbeen included.

Cardiac Status Final Rhythm Total________ ________ ________Sinus__A.F. _ _ _ _ _ _ _ _ _ _ _ _ _

Normal Heart 9 3 12Minimal dyspnoea, mild to 2 4 6moderate enlargement

Moderate dyspnoea, marked 1 4 5enlargement

Chronic cardiac failure, gross 1Ienlargement

Died, ? recurrent thyrotoxicosis 1Ipartially responsible

Died from unrelated causes 2 2 4Associated heart disease 4 1 5

predominantNot available for assessment, 1 3 4

but rhythm known -_-_-_-_iTotal 19 19 38

further men-tion, Case 3 because he illustratesthat A.F. may be the only clinical manifestationof persistent thyrotoxicosis and Case 4 becausehe was one of only two subjects that we haveever seen who developed recurrent thyrotoxico-sis after 1311 treatment.Case No. 3. A male aged 58 was admitted to

hospital wilth severe thyrotoxicosis, thyrotoxic myo-pathy and paroxysmal A.F. The thyroid was verylarge and as tracer investigation had shown a rapidturnover, 25 mc. 13'I were given in three divideddoses under quinidine cover, the pulse rhyythm re-maining regular throughout. He improved rapidly butthree months later he was found to have developedA.F. again. In the next two months A.F. persisted,this being the only clinical manifestation of persistentthyrotoxicosis but as the thyroid was still palpablea further 10 mc. 1'8I was given. One month laterhe was in sinus rhythm though he shortly developedpostradiation myxoedema.

Case No. 4. A man aged 71 was admitted toanother hospital in severe heart failure and A.F.,two years after a proven posterior myocardial infarct.When the heart failure had responded to treatmentthyrotoxicosis was diagnosed and 5 mc. 131I were givenwith a very satisfactory response, sinus rhythm beingrestored in two months; he remained euthyroid withno significant cardiac symptoms. 16 months later ata routine visit he was found to have developed A.F.again and to have lost some weight. Investigationsconfirmed the recurrence of thyrotoxicosis and an-other dose of "1'I was arranged but unfortunately hedied suddently before this could be given.

Cardiac Status after Completion of TreatmentThe cardiac status of these patients with

A.F. when they were euthyroid is summarisedin Talble 2. 18 were normal, or had only minimal

disability and these included all with paroxys-mal A.F. and the one patient who had A.F.transiently after the first dose of 131I. Althoughthere was a tendency for the patients withresidual A.F. to be less satisfactory, three ofthem had no symptoms, a normal-sized heartand an otherwise normal ECG, while threewho reverted to sinus rhythm had persistentmarked enlargement of the heart; two of thelatter had another cause for the heart disease.The importance of associated heart disease

in those patients with residual cardiac symptomsand signs was difficult to assess and only thosein whom such other causes were undoubtedhave been included. They were three withischaemic heart disease, two with rheumaticheart disease and one with severe hypertension.Only one of our subjects with A.F. had anginapectoris though there were several instances inpatients who had normal rhythm throughout.

Seven subjects, one of whom died from anindependent cause, had marked residual en-largement of the heart without clear evidenceof associated heart disease. Most of these wereelderly and in some instances it was probablethat thyrotoxicosis had been present for aconsiderable time. Except for Case 6 they hadfew symptoms, any exertional dyspnoea beinglittle more than would be expected for theirage though the lack of symptomns may havebeen partly due to the great improvementfollowing treatmenit of the thyrotoxicosis.Further they have shown no tendency to de-

by copyright. on June 12, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.41.481.663 on 1 Novem

ber 1965. Dow

nloaded from

November, 1965 STAFFURTH, GIBBERD and HILTON: Atrial Fibrillation

teriorate in the 1 to 5 years since becomingeuthyroid. Three patients had clinical andradiological evidence of mitral incompetence,but without an antecedent history of rheuma-tism or present evidence of mitral stenosis thismay be a legacy of thyratoxic heart disease.The patient who died, aged 73, did so froma refractory anaemia and at autopsy she hada very large heart, "surprisingly healthy" coron-ary arteries and small multiple areas of filbrosismicroscopically.Only two subjects had significant disability

and as they illustrate different aspects of theproblem they are briefly described:Case No. 5. A male aged 50 in 1959, when first

seen, had severe thyrotoxicosis for 18 months, A.F.for an indefinite time and heart failure for threemonths. He was difficult to treat with .31I for herequired 42 mc. in four separate doses over a periodof 16 months before he became completely euthyroid.In the next five years he continued to complainof exertional dyspnoea which interfered seriouslywith his work. He had moderate cardiac enlargementand an attempt has recently been made to reverthim to normal rhythm with the external D.C. de-fibrillator, unfortunately without success.

Case No. 6. A female, alged 61, presented at an-other hospital in September, 1962 with symptomsof progressive cardiac failure and A.F. for threemonths shown to be due to a toxic nodular goitre.As there was a poor response to carbimazole shewas referred and treated with 12 mc. "1"I in December,1962, and a further 10 mc. in July 1963 beforeshe became euthyroid. Slhe has severe persistentdyspnoea, a very large heart and repeated episodesof congestive cardiac failure. An attempt to reverther to normal rhythm with the DC defilbrillator hasbeen successful but it is too early to know whetherthis has been beneficial.

Cerebral Embolism in Association with A.F.and ThyrotoxicosisThere were two cases in the main series of

39 su'bjects with A.F. who had strokes pre-sumed to be due to cerebral emboli and inthe same period a further eight of our patientshave 'been treated surgically or withcarbimazole, none of whom have had a stroke.There have been no episodes of a cerebralvascular accident in the 131 subjects with sinusrhythm in the main series. The two cases whohad cerebral emboli are briefly described:

sCase No. 7. A female aged 56 in 1955 was admittedto hospital with a left hemiplegia from which shemade a good recovery. She had A.F. and cardiacfailure due to severe thyrotoxicosis with a largediffuse goitre and there was presumptive evidenceth'at the A.F. had developed two months before thehemi,plegia. She was treated satisfactorily withcaibimazole; in two months she reverted to sinusthythm and the drug was continued for two yearsuntil July 1957, sinus rhythm being present through-out. However six weeks after the carbimazole w-asstopped signs of thyrotoxicosis and A.F. returned.She was referred elsewhere for 13"I therapy, 10 mc.

being given in October 1957. There was only lapartial response to this dose and in November 1958she was again admitted to hospital with a poplitealemnbolism and severe signs of ithyrotoxicosis. Shewas given a further 8 mc. "'I in December 1958and 8 mc. in March, 1959 wlhioh led to normalthyroid function, though A.F. persisted. In April,1960 she had a further left hemiplegia, from whichshe has remained disabled though she was still alivein 1964.iCase No. 8. A female aged 77 in 1962, was

admitted to another hospital in heart failure withA.F. due to toxic nodular goitre and hypertension.She had probably had thyrotoxicosis for three yearsand A.F. for an unknown period. She was referredfor "31I treatment, 8 mc. 'being given in August, 1962with slight improvement in the next two months.She was admitted to a third hospital in January, 1963with a right hemiplegia from which she made apartial recovery. In August, 1963 she was againreferred from the first hospital with persistent thyro-'toxicosis and A.F. Another 8 mc. 131I were given withslow but steady improvement and in April, 1964 shewas euthyroid but A.F. was still present. However,three weeks later she was admitted under our carewith another right hemiplegia, the heart now beingin sinus rhythm. She made a moderate recovery butshe has a residual hemiplegia, hypertension and avery large heart.Of the 19 subjects with A.F. treated at St.

Thomas's Hospital in 1956-57, two died shortlyafter treatmenit, one from cardiac failure andone, who had previously had multiple pulmon-ary emboli, from a presumed cerebral embolusnow descri'bed.Case No. 9. A female aged 47 was admitted to

hospital in March, 1957 in severe cardiac failurewith A.F. due to a toxic nodular goitre, the goitrehaving been present for most of her life. In thepreceding four months there had been repeatedepisodes of sudden dyspnoea, with pleurisy, whichwere attrilbuted on good clinical grounds to recurrentpulmonary emboli, and she was put on anticoagulants,digoxin and mersalyl. Tihree weeks later 12 mc. 1311were given followed by Lugol's iodine and in twoweeks she was improved but cardiac failure andparoxysmal A.F. persisted. WThen the Lugol's iodinewas stopped she became frankly thyrotoxic withpersistent A.F. so a further dose of 131I was givenin July, 1957 at which time the anticoagulants werestopped.Three weeks later she had a sudden complete left

hemiplegia from which she soon died. The clinicaldiagnosis was presumed to be a cerebml embolusbut the post-mortem was un'helpful for no adequateexplanation of the hemiplegia was found. Therewas no evidence of cerdbral infarction nor of anyocclusion from throm'bus or emibolus in any branchof the carotid artery. Nor was there any evidence ofthe earlier pulmonary emboli, though these couldhave resolved. However, the heart was enlarged andantemortem thrombus was presenit in the tips ofboth atrial appendages and at the base of one ofthe papillary muscles. The coronary vessels werevery healthy, but microscopically there was someinterstitial fibrosis.Of the remaining 17 sulbjects, 10 (six with

persistent A.F.) are known to be alive andhave not had any strokes 'whilst seven (four with

667

by copyright. on June 12, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.41.481.663 on 1 Novem

ber 1965. Dow

nloaded from

POSTGRADUATE MEDICAL JOURNAL

persistent A.F.) have died, five from unrelatedcauses, one from a proven cerebral haemor-rhage and one, six years later, from a carcinomaof the thyroid which will be repotted elsewhere(Staffurth, 1965).A further patient has recently been seen

with a cere3bral embolus 11 years after treait-ment for thyrotoxicosis.

Case No. 10. A female aged 70 developed a suddenmild lefit hemiplegila followed two weeks later byan embolus in the right brachial artery for whiohshe was admitted to another hospital for anticoagu-lants. 11 years before she had been treated for'thyrotoxicosis and A.F. with methyl thiouracil whichhad been continued for two years. A.F. had persistedbut at no time had there been any suggestion of arelapse of thyrotoxicosis and subsequent investiga-tions confirmed that she was euthyroid.

DiscussionThe overall incidence of A.F. in thyrotoxico-

sis is about 10% (Means, DeGroot and Stan-bury, 1964), ithe 23% incidence in this seriesbeing due to the high average age of patientsreferred for 1311 therapy. Half of our patientswho originally had A.F. were in sinus rhythmwhen euthyroid, a figure which comparesfavouralbly with the reversion to normal rhythmof one-third of the sulbjects in a larger, butcomparable, series from Sheffield (Sandler andWilson, 1959). The difference in the two setsof results may be due to the number of caseswith paroxysmal A.F. as Sandler and Wilsondid not specifically mention paroxysmal. A.F.and 'they may not have included these cases.When our subjects with paroxysmal A.F. areexcluded the proportion ultimately in sinusrhythm was also only one-third.By comparison, a reversion rate of even

one-half compares unfavouralbly with someother series particularly when large numberwere treated surgically before 131,I was available.Dunhill, who was interested in this subject,first reported the success of thyroidectomy inpatients with irregular pulse rhythms in 1909,and in 1930 he was alble to report that 48 of100 cases with egtablished A.F. revertedspontaneouslv to normal rhvthm and another32 did so after quinidine, giving a total rever-sion rate of 80%, though this was with a 9%mortality.Hudson (1959) reported the results of thy-

roidectomy in 254 sutbjects with A.F., averageage 51.4 years, of whom 74% reverted sponitane-ously to normal rhythm after operation, themortality being only 2.4%. His series included61 with paroxysmal A.F. of whom 60 rever'ted,but more important 70% of those with estab-lished A.F. reverted to normal rhythm. He

commented that reversion was less likely whenA.F. had been present for more than a yearand when other forms of heart disease werepresent but even in this group there was areversion rate of one-third, that is, similar tothe results in all our patients with establishedA.F.On the other hand Delit, Silver, Yohalem

and Segal (1961) treated a large number ofthyrocardiacs with 1311 with results apparentlyas good as these surgical series. 226 (15%)of 1,492 patients had A.F. before treatmentwhilst this was present in only 51 (4%/O) of 1,294after treatment. Unfortunately no indicationwas given about the situation in the remaining198 patients. They also state that 107 of theircases had A.F. as the only manifestation ofheart disease, 80 out of 81 of whom" hadreverted to normal rhythm. This is so differentfrom our results, those of Sandler and Wilson(1959) and another large series from the U.S.A.(Chapman and Maloof, 1955) that there mustbe some unexplained reason. For instance, wereall their cases confirmed by ECG, as this is notspecifically stated? And what happened to the25% who were not followed up?From this brief review it would seem that the

results of surgical treatment are better thanwith 1311 though the material is not entirelycomparalble. Many of the subjects referred tous for 1311 are elderly with severe heart disease,their average age being more than ten yearsolder than those of Hudson (1959) and there isoften a suggestion that this is -the reason whythe physician asked us to trea't them. In otherinstances the same physicians refer youngerpatients with A.F. for thyroidectomy and thishas recently been our practice. These facts areprobably the main reason why our resultsare not as good as the surgical series butnevertheless statistical excuses must not beallowed to hide individual failures, and in thiscontext we' would particularly draw attentionto Case 7, where early surgery would un-doubtedly have been more successful, and alsoCases 5 and 6, where thyroidectomy wouldprobably have produced a better result if thethyrotoxicosis had been controlled earlier.The main deleterious effects from persistent

A.F., per se, are the risk of emibolus andpossible impairment of cardiac output. Cerebralembolism, leading to death or disabling hemi-plegia, is a common complication of A.F. dueto any cause bu't its occurrence in thyrotoxi-rosis has received little attention. One can neverbe sure in an individual paitient that a strokewas dtue to an embolism but the clinical his-

November, 1965668

by copyright. on June 12, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.41.481.663 on 1 Novem

ber 1965. Dow

nloaded from

November, 1965 STAFFURTH, GIBBERD and HILTON: Atrial Fibrillation

tories of our four subjec-ts (Cases 7-10),particularly the presence of emboli elsewherein three of them, suggest that these strokes weredue to cerebral emboli. No indication of thefrequency of this complication would be validfrom our data because this appraisal was partlyundertaken after our first two cases had beenseen and is thus biassed. However, the occur-rence of four cases is, in itself, an indicationthat a risk of embolism exists in patients withthyrotoxicosis and A.F.

Little attention has been paid to cerebralembolism in the recent literature on the manage-ment of thyrotoxicosis. However, it cannot beall that uncommon for Carter (1963) collectedsix instances from A.F. due to thyrotoxicosisout of 108 cases of cerebral emboli. Further,the occurrence of cerebral emibolism was re-cognised as a hazard, albeit a rare one, in thedays when large numbers of patients wijth A.F.were treated by partial thyroidectomy, whenthe emboli tended to occur at the time ofreversion to normal rhythm (Rundle, 1951;Hudson, 1959).

It is surprising that cerebral emboli have notbeen reported in patients who have beentreated with 1311 but it is possible that theyhave not been recognised. Chapman and Maloof(1955) reviewed 520 patients with 16 deathsin the follow-up, nine of whom were "aged,chronically ill and had vascular thrombosisin vital areas". Blomfield, Eckert, Fisher,Miller, Munro and Wilson (1955) had 30deaths out of 500 cases, in 18 of whom "deathwas due to cardiovascular disease presentingas congestive cardiac failure, myocardial in-farction or cerebral thrombosis", and Delitand others (1961) reviewing a large seriesstate that "the cardiac patients including 'pure'fibrillators, contribbuted a disproportionatelylarge percentage of the total deaths" withoutindication of the number or nature of thosedeaths. Were any of these "cerebral throm-boses" really emboli? That none of theseauthors have commented on the occurrenceor even the possibility of non-fatal emboli intheir cases which include a large number withA.F. is remarkable. A possible discrepancybetween these experiences and the earliersurgical ones is that 1311 being effective moreslowly, any untoward events are likely tooccur when patients are outside the immediatecare of the centre that arranged the treatmentand so may not be reported.

Although the heart may be involved in thy-rotoxicosis without the development of A.F.our patients in sinus rhythm have not been

considered here as the whole subject is wellreviewed by others (Cookson, 1959; Sandierand Wilson, 1959; Summers and Surtees, 1961;Delit and others, 1961). Our findings as regardsthe incidence and severity of heart disease aresimilar to those generally reported, in thatbefore treatment there is an increasing in-oidence of A.F., cardiac failure, and enlarge-ment of the heart with advancing age, and acorresponding decrease in a satisfactoryresponse to treatment. However, one can neverforecast how an individual patient will respond.We did not find, as have others, that thepresence of cardiac failure with A.F. beforetreatment adversely affected the chances ofreversion to normal rhythm and it should benoted that three of the four subjects over 75reverted to normal rhythm, one of themhaving had A.F. for two years (Case 8).

Sandler and Wilson (1959) and Summersand Surtees (1961) commented on the numberof subjects with residual heart disease, part-icu'larly enlargement of the heart, which couldnot be explained by associated disorders, andboth considered that there is a form ofpermanent heart disease which is a legacy ofthyrotoxicosis. Most of their subjects, likeseven of ours who were similar, were elderlyso that in an individual case it is impossibleto be sure whether or not an associated condi-tion, particularly ischaemic heart disease, isalso present. Sandler and Wilson (1959) re-viewed the literature on au'topsies on thissubject and added six of their own, five ofwhom had "healthy coronary arteries" andone, a woman aged 37 had patchy myocardialfiibrosis. This was the finding in our only avail-able autopsy, a patient aged 73 when she diedwith marked cardiac enlargement two yearsafter being euthyroid.Many of these patients appear to have had

undiagnosed thyrotoxicosis for a long time butSummers and Surtees (1961) reported animportant patient who developed thyrotoxicosisand marked cardiac enlargement in four monthswhile under their observation for anothercomplain't, and in whom the cardiomegalypersisted after treatment. In one of our sub-jects (Case 6) the antecedent history was onlya matter of months so we do not think onecan always blame delay in diagnosis for thisimportant complication.When these cases are considered together

they amount to a considerable number andthere is little doubt that a form of chronicheart disease may persist after the successfultrealtment of thyrotoxicosis. We suggest that

669

by copyright. on June 12, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.41.481.663 on 1 Novem

ber 1965. Dow

nloaded from

POSTGRADUATE MEDICAL JOURNAL

this be called post-thyrotoxic cardiomyo-pathy. The possibility of this developing is agood reason why all subjects with thyrotoxico-sis, particularly those with heart disease, shouldbe treated promptly and properly (Hudson,1959). Although l3sI therapy is a satisfactorymethod of treatment it should not be advisedautomatically in all sulbjects over 45, for therecent development of A.F., particularly in anotherwise healthy middle-aged subject, may bebetter treated by thyroidectomy, as the evidencesuggests that a return to normal rhythm is thenmore probable.

Surgical treatment also presents an opportun-ity to try to induce sinus rhythm, if this doesnot occur spontaneously, before the patientis discharged from hospital (Dunhill, 1930).After 31I Itherapy attempts at reversion to sinusrhythm have been less often reported thoughSandler and Wilson (1959) used quinidine ona small num-ber with some success. We haverarely used quinidine because of practicaldifficulties, for a-fter L311I treatmenit one mustwait until the patient is euthyroid and almos-tcertainly back at work before an attempt atreversion is likely to succeed. The patient, nowwithout symptoms, has to be readmitted forseveral days, wh'ich is psychologically badmanagement. With the advent of DC defibril-lation it is possible that more attempts will bemade to revert patients wiith persisten-t A.F. andwe have now itried this on itwo subjects (Cases5 and 6).

Because of the de1ay in the control ofthyrotoxicosis 'that may occur after the firstdose of 1311, a-ttempts have been made to pre-cede or follow the '%1I with antithyroid drugs inorder to ensure early control. This method hasits disadvantages for Crooks, Buchanan, Wayneand Macdonald, 1(1960) found that methylthiouracil given ibefore 1the 131I reduced con-siderably 'the number of patients who wFrecured by the first dose and also that the average-total dose for cure was increased. Bloomfieldand others (1959), with a smaller number ofcases, did not notice any such differences. Ourexperience with carqbimazole has tended toagree with ithat of Crooks and others, but ourseries has not been controlled and we havethought that the less good response may havebeen due to seledtion, for it is ithe more severecases with 'large thyroids that have more oftenbeen pretrealted in this way.Crooks and others (1960) did not discuss an

impoftant difficulty that occurs in -the manage-menst of -the patients Who did not respond to thefirst dose of 3111, namely that such a patient,

who has become normal on antithyroid drugsat the time of administraltion of the I31J, at somestage has to stop the drug to see if the 131J hasbeen effective. Those who have not been cu.redby ithe -first dose may ithen become thyrotoxicagain, having 'been controlled for severalmon(ths, before their final cure. This is anuntidy method of management tha't calls forclose supervision, the difficulties being wellillustrated in our Case 3.Werner (1962) and o;thers have tried begin-

ning the antithyroid drug after the first dose of13'I, a method we 'have 'been using in seledtedcases recently. Our experience is, as yet, tooshort to assess i-ts effectiveness 'but it is clearthat the same difficulties in management exist.Nevertheless, it does ensure Ithat mogt patientsbecome euthyroid rapidly and we 'think tha:t thedisadvantages are preferable to the possibilityof otherwise uncontrolled thyrotoxicosis. Treat-menit with an anitithy.roid drug, whether beforeor after the I3II should probably be undertakenin all patients with severe thyrotoxicosis,selected for definitive 'treatment With 184I, par-ticularly if 'there is A.F. or other evidence ofheart disease.We agree with Sandler and Wilson (1959) that

131J therapy is usually satisfactory for thyro-cardiacs and that it should be used in severelyill or elderly patients because of the increasedoperative mortality in this group '(Crile, 1949).However, 'it is not always immediately success-ful for a small number of subjects may getlittle benefit from 'the firs't dose of 131. Also,patients with estalblished A.F. even when theyhave improved considerably, may not revertto normal rhythm until they are unequivocallyeuthyroid '(Maloof and Chapman, 1951), a factwell illustrated by Case 5. As a consequence ofeither of these factors, A.F. may persist for acongiderable 'time so that the ultimate prospectof reversion is seriously lessened.Some of ithe practical difficulties of ""I

therapy have been stressed because they havebeen insufficiently discussed in -the past and alsobecause they have been particularly evidentin these patients with A.F. Further experienceWill be necessary to see whether -the use of anantithyroid drug wi'th '3I will (be more success-ful and whether this combination will be aseffective as thyroidectomy.

Summary1. An assessment has been made of the resultsof radio-iodine treatment in 39 patients withthyrotoxicosis and atiial fibrillation (A.F.).Nineteen similar patients treated elsewhere have

670 November, 1965by copyright.

on June 12, 2020 by guest. Protected

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.41.481.663 on 1 N

ovember 1965. D

ownloaded from

November, 1965 STAFFURTH, GIBBERD and HILTON: Atrial Fibrillation 671

been followed up with particular reference tothe possible occurrence of cerebral embolism.2. After completion of treatment half of thepatients were in sinus rhythm 'but when thosewith paroxysmal A.F. were excluded the pro-portion was reduced to one third. Theseresults compare unfavourably with earlier sur-gical experiences and possible reasons for thisare discussed.3. The cardiac status after treatment is des-crilbed and attention drawn to a persistent formof heart disease which is not uncommon andwhich it is suggested should be called post-'thyrotoxic cardiomyopathy.4. Four instances of -suspected cerebral emboliare descrilbed and attention drawn to the pos-sibility of this complication in thyrotoxicosiswith A.F.5. The use of ,antithyroid drugs before orafter radioiodine is discussed with emphasis onthe necessity for care in management that thisentails.6. It is concluded that all subjects with thyro-toxicosis and A.F. should be itreated ener-getically 'whatever method is chosen.We thank the numerous physicians who have

referred the patients to us and to Dr. J. A. C. Flemingfor permission to include patients treated at St.Thomas's Hospital. We are grateful to Dr. R. L. R.Hartley for attempting D.C. defibrillation on twosubjects. We are particularly grateful to Mr. J.Whale for technical assistance. J.S.S. was in receiptof a personal part-time grant from the MedicalResearch Council.

REFERENCESBLOMFIELD, G. W., ECKERT, H., FISHER, M., MILLER,

H., MUNRO, D. S. and WILSON, G. M., (1959):Treatment of Thyrotoxicosis with '31I, Brit. med. J.,i, 63.

CARTER, A. B., (1963): Strokes, Proc. roy Soc. Med.,56, 483.

CHAPMAN, E. M. and MALOOF, F., (1955): The Use ofRadioactive Iodine in the Diagnosis and Treatmentof Hyperthyroidism, Medicine (Baltimore), 34, 261.

COOKSON, H., (1959): The Thyroid and the Heart,Brit. med. J., i, 254.

CRILE, G. (1949): Thyroid Disease, Philadelphia andLondon: J. B. Saunders.

CROOKS, J., BUCHANAN, W. W., WAYNE, E. J. andMACDONALD, E. (1960): Effect of Pretreatment withMethylthiouracil on Results of "3'I Therapy, Brit.med. J., i, 151.

DELIT, C., SILVER, S., YOHALEM, S. B. and SEGAL,R. L. (1961): Thyrocardiac Disease and itsManagement with Radioactive Iodine, J. Amer.med. Ass., 176, 262.

DUNHILL, T. P., (1909): Remarks on PartialThyroidectomy, Brit. med. J., i, 1222.

DUNHILL, T. P. (1930): Auricular Fibrillation inThyrotoxic Conditions, Brit. J. Surg., 17, 424.

GREEN, M. and WILSON, G. M., (1964): Thyrotoxi-cosis treated by Surgery or Iodine-1 31. Withspecial reference to Development of Hypo-thyroidism, Brit. med. J., i, 1005.

HUDSON, R. V., (1959): Toxic Goitre and AbnormalCardiac Rhythm, Proc. roy. Soc. Med., 52, 810.

MACGREGOR, A. G. (1957): Simplified RadioactiveIodine Therapy, Brit. med. J., i, 492.

MALOOF, F., and CHAPMAN, E. M. (1951): Radio-iodine Therapy in Hyperthyroidism: CardiacProblems, J. clin. Endocr., 11, 1296.

MEANS, J. H., DEGRooT, L. J. and STANBURY, J. B.(1964): The Thyroid and its Diseases, New Yorkand London: McGraw-Hill.

RUNDLE, F. E. ,(1951): Joll's Diseases of the Thyroid,London: William Heinemann.

SANDLER, G. and WILSON, G. M. ,(1959): The Natureand Prognosis of Heart Disease in Thyrotoxicosis,Quart. J. Med., 28, 347.

STAFFURTH, J. S. (1965): Carcinoma of Thyroidafter Radioiodine Treatment in Thyrotoxicosis, Brit.J. Radiol. To be published.

SUMMERS, V. K. and SURTEES, S. J., (1961): Thyro-toxicosis and Heart Disease, Acta med. Scand.,169, 661.

WERNER, S. C., (1962): The Thyroid, New York:Harper and Row.

by copyright. on June 12, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.41.481.663 on 1 Novem

ber 1965. Dow

nloaded from