HASHIMOTO'S DISEASE - pmj.bmj.com · i85 HASHIMOTO'S DISEASE: A REVIEW By DOUGLAS R. K. REID, M.D., F.R.C.S.* ' If wesummarise the present clinical findings, wehadfourpatients, all

i85

HASHIMOTO'S DISEASE:A REVIEW

By DOUGLAS R. K. REID, M.D., F.R.C.S.*

' If we summarise the present clinical findings,we had four patients, all beyond forty years of age,not living in a goitre district, and with a familyhistory negative for goitre. . . . They had neverbeen seriously ill, and were without any of theserious infectious diseases frequently followed bydisease of the thyroid, especially strumitis. Luesand tuberculosis were excluded clinically. Thethyroid enlargement was discovered incidentallyin all cases.

' We found a not very rapid enlargement of thethyroid, with good general health.... Both lobeswere involved in all cases.... Severe symptoms ofpressure such as dyspnoea and aphonia have neverbeen present.... Not a trace of inflammation ofthe thyroid could be found. The mass was freelymovable, but the consistency was usually so firmthat one would think of malignancy or Riedel'sstrumitis. There was neither regional nor generallymphadenopathy. No other visceral abnormalitywas found.'The operative findings did not demonstrate

much adherence to adjacent structures, as incases of Riedel's thyroiditis... . Post-operativehoarseness was present in all cases, though slightin amount. In the post-operative course, anoedema of the entire body is said to have occurred,disappearing after the internal administration ofthyroid....

' The chief findings in all cases were:' (i) Formation of numerous lymph follicles.' (ii) Striking changes in the epithelial tissue of

the acini (atrophy) of the thyroid andtheir contents (absence of colloid).

'(iii) Extensive proliferation of connectivetissue and diffuse round-cell infiltration(lymphocytes and plasma cells).'-

Hashimoto, I9I2, quoted by Lee and McGrath,'937.The disease described by Hashimoto in these

terms nearly 40 years ago has ever since beenattended by confusion and controversy. Its veryexistence as a clinical and pathological entity hasbeen alternately affirmed and denied. It has beenrelated to thyrotoxicosis on the one hand, and tomyxoedema on the other. It is not fatal, yet com-

* From -the Department of Surgery, PostgraduateMedical School of London, Hammersmith Hospital.

paratively little is known of its course whenuntreated. Its diagnosis rests on histological find-ings, yet no one of these is pathognomonic, and theultimate criterion is not the quality of the changesinvolved but their extent and quantity. It has beengiven at least three names and a score of aetiologies,and none has yet been found satisfactory. Andthough our knowledge of normal and disorderedthyroid function has advanced so dramatically inthe last decade, still no clear light has been shed onthe pathogenesis of Hashimoto's disease. Itappears to have no analogue in any other organ ordisease process, yet neither has it been integratedwith the peculiar physiology and pathology of thethyroid.No further definition of the disease, other than

the quotation above, will be offered at present;yet we must not too readily assume that Hashi-moto's description, based as it was on only fourcases, is necessarily final and completely com-prehensive. The purpose of this paper is toexamine the literature and to attempt to indicatewhat in retrospect seems paradoxical, and whatin prospect may prove significant and of value inour future understanding of the disease. Furtherdefinition must await further elucidation,

IncidenceMost cases of Hashimoto's disease have been

reported from Britain, the United States andJapan. It occurs on the Continent, but accordingto Joll (I939) is relatively rare in the endemicgoitre areas of the Tyrol and Switzerland; myown small personal experience would suggest thatit is similarly uncommon in New Zealand.

It is by no means a common condition, and hasbeen variously reported from British and Americansources as constituting from 0.17 to 1.7 per cent.of all thyroid glands removed at operation(Table i). Graham in I931 reviewed o04 casesof thyroiditis, and could find only 24 which, in hisopinion, conformed to Hashimoto's description.McClintock and Wright (1937) accepted 50 cases,and McSwain and Moore (I943) 71 cases, whilein 1948 Goldberg and Davson referred to over200 case reports in the literature. The largestindividual series has appeared since then, andincluded 78 cases from 25,ooo thyroidectomies, anincidence of 0.31 per cent. (Marshall et al., i948).

copyright. on 7 M

arch 2019 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.27.306.185 on 1 A

pril 1951. Dow

nloaded from

http://pmj.bmj.com/

i86 POSTGRADUATE MEDICAL JOURNAL April 1951

TABLE I

SHOWING THE INCIDENCE OF HASHIMOTO'S DISEASE RELATIVE TO THE NUMBER OF THYROID GLANDS REMOVED

Number of Number of cases ofAuthor Year thyroidectomies Hashimoto's disease Incidencei

Lee (U.S.) .. .. . .. . 935 i,8oo 3 0. I 7%Renton et al. (G.B.) .. .. .. .. 1938 48 7 1.7 %,Joll (G.B.) . 1939 5,650 5 0.9 oGraham (G.B.). .. .. .. 939 670 6 o.9 oiKeynes (G.B.) .. .. .. 939 I,6oo 25 I.56%McSwain and Moore (U.S.) 1943 1,999 15 0.75%/Ficarra (U.S.) .. .. .. .. 1946 1,938 9 0.4 %Marshall et al. (U.S.) .. .. .. I948 25,000 78 0.31 ,Crile (U.S.) .. .. .. 1948 900 3 .33Lasser and Grayzel (U.S.) .. .. 949 1,595 9 0.49%

Of the 28 cases which Davison and Letton (I949)observed between 1935 and 1949, no fewer than26.had occurred within the last six years of thatperiod.

It would be unwise to assume that Hashimoto'sdisease is becoming any commoner, but Table 2shows a peculiar alteration in the numbers of casesreported, relative to those of Riedel's thyroiditis.Up to the early 'thirties Riedel's disease wasmore common than Hashimoto's, the ratio being2: i (Graham, 1931), or 3: i (Lee, 1935). By 1937the ratio was nearly I: I (McClintock and Wright),and in Joll's series (I939) it was completelyreversed and was now I: 8. Paterson and Starkey(1948) reported a ratio of i: i, as did, roughly,Lasser and Grayzel (I949), but in Marshall's largeseries it was i: 1. It is true that Marshall appliedthe term 'struma lymphomatosa' to mild,moderate and marked instances of diffuse lym-phoid infiltration, but limited the use of Riedel'seponym to the cases of marked fibrosis. Com-paring only the fully developed cases in this series,the ratio becomes i: 4 in favour of Hashimoto'sdisease.

. Te variability of the ratio and its apparentcomplete reversal in the course of 20 years maybe accounted for in several wavs:

(a) It may be statistical error, due to the smallnumbers involved.

(b) It may be due to an actual fall in theincidence of Riedel's disease.

(c) It may reflect difficulty, and perhaps chang-ing opinions, in the pathological classification ofcases of thyroiditis. In this connection, it mustbe emphasized that the authors whose figures arequoted are those who have most sharply distin-guished Hashimoto's from Riedel's disease, andat least to that extent have shared similar criteria.

(d) The alteration in the ratio has coincidedwith the perfecting, and consequent popularity,of the operation of thyroidectomy; it is feasiblethat, since fewer patients nowadays are content toexist goitrous for long, treatment of the earlierstage of a disease process has reduced the incidenceof its later stage..Considerations of age and sex incidence have

played a large part in the distinction betweenHashimoto's and Riedel's diseases. It is generallyagreed that both are more common in femalesthan in males; this seems to be particularly trueof Hashimoto's disease and, according to Davisonand Letton (1949), only eight male cases had beenreported at that time.Age incidences yield more room for criticism

TABLE 2

SHOWING THE INCIDENCEOF HASHIMOTO'S DISEASE RELATIVE TO THAT OF RIEDEL'S AND OTHER FORMS OF THYROIDITIS

No. of cases of No. of cases of No. of cases of otherAuthor Year Hashimoto's disease Riedel's disease forms of thyroiditis

Graham (U.S.) .... 931 24 4I 39Eisen- (U.S.) .. .. . 934 3 4Lee (U.S.) .. .. .. 1935 26 90McClintock and Wright (U.S.) .. 1937 50 60oJoll (G.B.) ... . . 1939 8i 0oPaterson and Starkey (U.S.) .. I948 II I 2Marshall et al. (U.S.) .. 1 948 78 7 i02Means (U.S.) .. ... .. 1948 12 o 3Crile (U.S.) .... . .. 1948 14 I 27Lasser and Grayzel (U.S.) .. 949 9 7 8

copyright. on 7 M


http://pmj.bm

j.com/

Postgrad M


pril 1951. Dow

nloaded from

http://pmj.bmj.com/

April 1951 REID: Hashimoto's Disease 187

TABLE 3SHOWING THE AVERAGE AGE INCIDENCE OF CASES OF HASHIMOTO'S AND RIEDEL'S DISEASES IN VARIOUS SERIES

Hashimoto's disease Riedel's disease

Author Year Number Average age Number Average age(years) (years)

Graham .. .. .. .. .. 1931 24 52.4 4 36.2Lee . .. .. .. .. .. 935 26 476 43.6Clute et al. .. ..... 1935 9 555 -

McClintock and Wright .. .. . . 1937 50 49.2 6o 40.3Joll .. .. .. .. . 1.939 8i 57.6 5 44.4McSwain and Moore .. .. . 943 71 47Schilling 1... .. .. 945 5 30.4Parmley and Hellwig .. . . .. 1946 14 44.7Paterson and Starkey .. .. .. 1948 I 43.8 II 44.8Marshall et al . . .. .. 1948 78 50 7 46Crile . .. .. .. .. 1948 14 49 I 51Davison and Hellwig .. .. .. I949 28 377 -Lasser and Grayzel .. .. .. 1949 9 38 7 46

and speculation (see Table 3). Graham (1931)reported an average age incidence of 52.4 yearsfor Hashimoto's disease and 36.2 years for Riedel's.Presenting similar figures, Joll argued that the twoconditions could not therefore represent stages ofa single process. If later figures show anything,however, it is that the average ages for the twoconditions have tended to approximate, until thereis little apparent difference. It is doubtfulwhether, in fact, these average-age figures are ofmuch value, either in describing the diseases or indistinguishing between them, as the range coveredis so great. It is certainly true that most cases ofHashimoto's disease occur after the age of forty,but, as Table 4 shows, it is not rare in youngerpeople. Many of the recorded cases have beenbelow the age of thirty, and examples belowtwenty have been reported by Bogart (1928), Joll(1939), Schilling (I945), Parmley and Hellwig(1946), Means (I948) and Davison and Letton(I949). The youngest case on record seems to

be a six-year-old female reported by Lasser andGrayzel (I949).

Joll (I939) stated that Riedel's disease occurredmostly below the age of forty, but Crile (I948) hasrecently stated that it is found most typically afterthe age of fifty. And, on the other hand, Merring-ton (1948) has pointed out that Schilling (I945)described Hashimoto's disease as occurring be-tween the ages of forty and sixty, and yet in thesame paper reported six cases, of which the agesof five are given as respectively seventeen, twenty-seven, twenty-eight, thirty and fifty-two years. Itis clear that a consideration of the literature hardlybears out the contention of Graham (193 ) andJoll (I939) that the age factor in itself forms a firmdividing line between the Hashimoto and Riedeltypes of thyroiditis.Clinical FeaturesThe most common symptom is the develop-

ment of a goitre. Its average weight, according to

TABLE 4

SHOWING THE AGE DISTRIBUTION OF HASHIMOTO'S DISEASE

Number of cases

Author Year Age Total10-20 20-30 30-40 40-50 50-60 60-70 70+

Lee .. .. . .. 1935 2 5 8 6 3 2 26McSwain and Moore .. 1943 2 7 12 19 20 7 4 71Means .. .. .. 948 I 4 2 3 2 12

TOTAL .. .. 3 9 21 29 29 12 6 109

SHOWING THE AGE DISTRIBUTION OF RIEDEL'S DISEASE (LEE'S SERIES, 1935)Age

10-20 20-30 30-40 40-50 50-60 60-70 70+Number of cases .. 0- 24 28 14 8 i

copyright. on 7 M


http://pmj.bm

j.com/

Postgrad M


pril 1951. Dow

nloaded from

http://pmj.bmj.com/

POSTGRADUATE MEDICAL JOURNAL

McSwain and Moore (i943), is 91 gm., and,according to Marshall (1948), o00 gm. Jaffe (I937)states that he has frequently found at autopsyglands of normal size or less which displayed thehistological characteristics of Hashimoto's disease.The goitre may grow slowly, or it may appar-

ently attain its maximum size within a year; itsaverage duration, according to Joll, is 4.2 years.There is rarely evidence of previous thyroid en-largement and, untreated, the condition shows notendency to regress, as Crile (1948) has noted ina patient whose goitre was present for eight yearsbefore operation.Symptoms referable to the goitre may or may

not be present. Their average duration is givenby McClintock and Wright (1937) as 1.3 yearsand by Marshall (1948) as 26 months. Pain andtenderness are generally reported to be uncom-mon, but are noted by Paterson and Starkey (1948)in 5 out of their i cases and'by Crile (1948) in 3out of I4 cases. Pressure symptoms are on thewhole slight, and seem to be proportionate to thesize of the goitre. X-rays of the trachea frequentlyshow deviation, lateral or antero-posterior com-pression, or a forward displacement which is dueto encirclement of the trachea by the thyroidlobes.

Typically, the goitre is a diffuse enlargement,involving both lateral lobes, isthmus and pyra-midal lobe. The right lobe is often the larger,and this is said by Joll to reflect the normal asym-metry of the gland. Unilateral cases have occurred(Means, 1948) and, in others, although the histo-logical changes are uniform throughout both lobes,one of these may be disproportionately large.I have recently, for example, removed a gland ofwhich one lobe (the right) weighed i8o gm. andthe other only 40 gm.The goitre is of firm, rubbery consistence, and

its surface is smooth, lobular rather than nodular.It is usually mobile, and apparently free ofadhesions to the overlying muscles. The cervicallymph-nodes are not, as a rule, palpable, althoughthe so-called 'Delphian' node (Cope, 1948),which lies above the isthmus, is often observed atoperation.

Usually the onset of Hashimoto's disease isinsidious, but fever and leucocytosis have beenobserved (Means, I948). Pedigo and Abramson(I947) describe a case in which the onset wassudden and was accompanied by fever, painfuljoint swellings and evidence of thyrotoxicosis;the histological sections were passed by Foot asbeing characteristic of Hashimoto's disease.The patient may be euthyroid when first seen,

or hypothyroid as was the case in 8 per cent. ofMarshall's series (1948), but many authors claimto have observed an early hyperthyroid phase.

Thus Davison and Letton (1949) refer to emotionalupset and nervousness; McSwain and Moore(i943) state that loss of weight is more commonthan gain initially; Parmley and Hellwig (1946)mention tachycardia, sweating and tremor insome cases, while Hori (I946) and Sinaiko et al.(1948) report mild exophthalmos in patients withHashimoto's disease.

Marshall (1948) believes that alteration of thegeneral health is not a prominent feature, unlesshypothyroidism is present, and states that suchsymptoms as nervousness and fatigue are observedmostly in patients with 'a strong tendency toneurosis or a phobia of cancer.' Schilling (1945)considered that.mild cervical pressure, rather thanthyrotoxicosis, was the cause of nervousness,vertigo, tremor and insomnia. The association offrank clinical thyrotoxicosis with the histologicalfeatures of Hashimoto's disease has, however, beenreported by several authors, including Polowe(1934) and Eden and Trotter (I942); further,basal metabolic rates have been recorded up to+57 per cent. (Goldberg and Davson, I948),+63 per cent. (McSwain and Moore, I943) and+69 per cent. (Pedigo and Abramson, I947).

Joll accepted the existence of a hyperthyroidphase as an accompanying and perhaps incidentalfeature of the early stage in some cases, but hedenied the theory which had been propounded byEason (1928) and supported by Vaux (1938), thatHashimoto's disease was the end-result of a burnt-out thyrotoxicosis. It is clear that there is a seriouslack of clinical agreement as regards the associationof toxic features with Hashimoto's disease, butthere is no doubt that its course is towards pro-gressive hypothyroidism in almost every case.Operation merely serves to hasten its onset, andMarshall found that 75 per cent. of his casesshowed post-operative hypothyroidism. Gener-ally this may be controlled by thyroid medication,but some cases, according to Crile (1948), do notrespond satisfactorily. He prefers the term'hypometabolism' to describe their condition,and emphasizes systemic features such as anaemia,achlorhydria and a lack of well-being, which maylast for years after operation, in spite of thyroidadministration.

Recurrence of lymphadenoid goitre after opera-tion does not seem to occur, except where theoriginal operation has been a unilateral lobectomyand 'recurrence' takes place in the remaininglobe.By contrast, the clinical features of Riedel's

thyroiditis are listed by Joll (I939)as a relativelymore rapid development of a smaller, and initiallyunilateral, goitre, which produces marked pres-sure symptoms in a large proportion of patients,and which is not associated with pre-operative

x88 April i95xcopyright.

on 7 March 2019 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.27.306.185 on 1 April 1951. D

ownloaded from

http://pmj.bmj.com/


':.

~ ~ 0

X..

A-o

co

CA *i-; ;-;i;~"~';'

00

,'"" E " f CrC CO

*,, · olE; i~

·.'.. ..., , ,

Cao

...~J1.. . . ~j ·j:..· .? .:......:.

CO

oC

I.

6. 0o.....·....

V11

~:~'i~"·!":~'':~ :';'":... ·s

c co

s- 0 k." l

Ca"':

1~ ,.i*~E:

·aCa·SBl .:. ' ,..-.,.':::'::':':.".~x'...

:...

·I

Cddgo

o.- d

G1

copyright. on 7 M


http://pmj.bm

j.com/

Postgrad M


pril 1951. Dow

nloaded from

http://pmj.bmj.com/


hypothyroidism so often as is Hashimoto's disease.It is doubtful, however, whether many of thesedifferential points are entirely valid. Thus, judg-ing from Marshall's figures, there appears to belittle difference between the duration of the twoconditions, and Crile (I948) has specificallydescribed the goitre in Riedel's disease as a' bulky' tumour. Further, while Marshall found8 per cent. of his cases of' struma lymphomatosa '

hypothyroid before operation and 75 per cent.afterwards, his corresponding figures for fibrousthyroiditis were 7 per cent. and 27 per cent. Inthe histologically advanced cases, post-operativehypothyroidism was observed in respectively 93per cent. and 57 per cent. of the lymphoid andfibrous types. The difference before operationat least is not appreciable. It is clear, althoughit is not surprising, that the fibrosing conditiondescribed by Riedel does produce more strikingpressure effects than does lymphadenoid goitre,but as Merrington (I948) has stated, '. .. thereis a single broad clinical picture common to both(diseases), and attempts to draw a line betweenthem have not been convincing.'Pathology

It may ill become one who is not speciallytrained in pathology to discuss the morbid changesin a subject so complex and so difficult of inter-pretation as Hashimoto's disease. A general dis-cussion based on the literature may not, however,be considered presumptuous.The lymphadenoid goitre is commonly of

moderate size; the enlargement may be diffuse,asymmetrical or infrequently unilateral. Its sur-face is lobulated and its consistency firm. Whencut, the tissue has been described in terms rangingfrom friable to tough, and its vascularity is asa rule reduced. In colour it is white, grey, pinkor ' lavender yellow' (Davison and Letton, I949).The cut surface is homogeneous; colloid is absentor scanty, and there is a noticeable absence ofnecrosis, haemorrhage or calcification. Thecapsule may be moderately thickened.Graham (193i) and Joll (I939) have regarded

the absence, in Hashimoto's disease, of extra-glandular infiltration and adhesions to surroundingstructures as a cardinal point in its differentiationfrom Riedel's thyroiditis. Eisen (I934) has, how-ever, criticized Graham's paper on the score thathe failed to mention such changes in cases wherethe original authors (e.g. Shaw and Smith, I925)specifically described them. The same criticismmust be made of Joll's reference to Renton,Charteris and Heggie's paper (1938); he regardedtheir cases as examples of Hashimoto's disease,but ignored the fact that they described and illus-trated lymphoid and fibrous infiltration of the

sternothyroid and sternomastoid muscles in theirfirst case. Taylor (I95i) has recently seen a caseof this type, in which extensive infiltration ofmuscles by the lymphadenoid process has beendemonstrated histologically.

Eisen's own cases showed ' definite extra-capsular adhesions,' as also did those reported byOldfield (I948) and Merrington (I948). Hori(I946) stated that in all of his cases-two ofRiedel's and three of Hashimoto's disease---thegland was adherent to the trachea and difficult toremove. Indeed, Sinaiko (I948) found in onecase that the thyroid was ' so extremely adherentto the trachea, that no plane of dissection couldbe found.' These cases are, no doubt, isolatedinstances, but they seem to occur sufficientlyfrequently to warrant consideration, and perhapsto cast a shadow of doubt on yet another ofthe accepted criteria of distinction betweenHashimoto's and Riedel's diseases.

HistologyThe histological picture comprises diffuse

lymphocytic infiltration, more or less numerouslymph follicles, minimal fibrosis and certainepithelial changes.

Throughout the gland an abundance of lympho-cytes infiltrates the peri-acinar and peri-lobulartissues; focal lymphoid aggregates form bothpseudo-follicles and true germinal centres, whichcontain reticulo-endothelial cells, plasma cells andlymphocytes. Monocytes are found in smallnumbers, but polymorphs are usually absent.Foreign-body giant cells are present, and also'false' giant cells which appear to consist ofacinar cells lying in colloid and to represent theremnants of a disintegrating acinus.

Fibrous tissue is present in variable amount;classically, it is minimal in amount, fine in texture,and arranged in whorls around lobules of thyroidtissue. Schilling (I945) admits that there may becomplete fibrous replacement of the lobule, butdoes not accept any relationship with Riedel'sdisease.

Joll regarded these changes as being primary,and pathognomonic only if they were diffuse,uniform, and included true germinal centres. Itseems desirable at this point, however, to devotesome consideration to histologically similar changesof a localized nature. Simmonds, in 19I3, foundfocal lymphoid changes in 5 per cent. of normalthyroids, in 15 per cent. of non-toxic goitres andin 75 per cent. of toxic goitres. In the last cate-gory, his figure was confirmed and indeed exceededby Kocher, Wegelin and Hertz (I943). Thesechanges were found predominantly in females,and only on subjects over the age of thirty.Graham (193I) recognized these findings and

April 1 95 1I90

copyright. on 7 M


http://pmj.bm

j.com/

Postgrad M


pril 1951. Dow

nloaded from

http://pmj.bmj.com/

REID: Hashimoto's Disease

stated, rather optimistically, ' . . . were it possibleto assume an antecedent hyperthyroidism, anexplanation of the pathogenesis of struma lympho-matosa would not be difficult.' Joll, on the otherhand, admitted the existence of a hyperthyroidphase in somL cases, but would not accept focallymphoid changes as being of the same nature asHashimoto's disease.

Hellwig (1938) found lymphoid foci in 23.3per cent. of cases of exophthalmic goitre. Vaux(1938) and Goldberg and Davson (1948) havetraced lymphoid changes along a scale rangingfrom hyperplastic goitre on the one hand, toHashimoto's disease on the other, and believedthat the degree of clinical toxicity was inverselyproportional to the extent of lymphoid involve-ment. Recently Greene (I950) has distinguishedbetween simple lymphocytic infiltration and'lymphadenoid foci,' containing true germinalcentres, and has found the latter in 44 out of I6Ithyroids removed for toxic symptoms. These44 cases were all females, ranging in age fromeighteen to seventy-five. Two were examples ofHashimoto's disease, and out of 43, no fewer than20 developed ' frank and definite hypothyroidism 'post-operatively. By contrast, hypothyroidismdid not develop in any one of those II7 cases inwhich no lymphadenoid foci were found. Greenebelieves that such foci are not merely histologicallysimilar to Hashimoto's disease, but that in factthey constitute its earliest stage. Whitesell andBlack (1949), at the Mayo Clinic, have studied 86cases of exophthalmic goitre in which lymphocyticinfiltration and fibrosis were present in the glands.Their conclusions are:

(a) Lymphocytes and fibrous tissue are an in-tegral part of the histological picture ofexophthalmic goitre.

(b) Thyroid acini may be replaced by lympho-cytes alone, or by both fibrous tissue andlymphocytes.

(c) Patients with the mixed fibro-lymphocyticchanges were somewhat older thanpatients with lymphocytic type.

(d) Female patients showed greater replacementthan male patients in each decade.

(e) With greater degrees of replacement thebasal metabolic rates were lower, and theincidence of auricular fibrillation less.

(f) The incidence of exophthalmos increasedprogressively with increasing degrees ofreplacement.

(g) The incidence of post-operative myxoedemarose progressively with increasing degreesof replacement. In some cases showing40 to 50 per cent. replacement, theincidence of post-operative myxoedemaapproached 70 per cent.

It is thus evident that a strong case can be madeout for linking focal lymphadenoid changes intoxic goitres with Hashimoto's disease. Ingeneral terms, it might be said that these changesare, in some cases of hyperplastic goitre, theprelude to and the accompaniment of diminishingendocrine function.The lymphadenoid features of Hashimoto's

disease are so striking that they have tended toovershadow those changes in the acinar epitheliumwhich Hashimoto himself recognized and de-scribed as degeneration and atrophy. Joll (I939)and Schilling (1945) have described these changesvery fully. The acini are small and shrunken,colloid is scanty or absent, and there are occasionalintra-acinar collections, apparently of desquamatedepithelial cells or even lymphocytes admixed withresidual colloid. According to Schilling, thepicture is one of diffuse and uniform epithelialdegeneration, but Joll recognized the occurrenceof epithelial hyperplasia. In this he was inagreement with Shaw and Smith (1925), William-son and Pearse (1928) and many other authorsand, like them, believed that it was compensatoryhyperplasia following strangulation and destruc-tion of thyroid tissue. Renton et al. (1938) alsodescribed hyperplastic changes, but found themdifficult to reconcile with clinical hypothyroidism.An attempt to measure the relative quantity of

lymphoid and epithelial tissue in lymphadenoidgoitres has been made by Parmley and Hellwig(1946), and their results, if substantiated, are ofthe greatest possible significance. They foundthat only one-third of the total thyroid enlarge-ment could be attributed to lymphoid tissue; theremaining two-thirds consisted of epithelial tissue,indicating a very considerable increase in its totalamount.

Moreover, most authors have noted the extensivereplacement of normal epithelial cells by large,pale, eosinophilic cells. These may occur singly,or more usually in small but complete acini whichare devoid of colloid. In some cases the cellsmay be so packed together that they appear assolid areas, such as Renton et al. have described as'syncytial hyperplasia.' Hellwig (1938) relatedthe cells to a similar type seen in chronic mastitis,but Womach (i944) identified them with the so-called Hiirthle type of cell, from which indeed theyare indistinguishable. So characteristic are they ofHashimoto's disease, that Crile (1948) and Mar-shall (1948) regarded ' acidophilic degenerationas an integral part of the histological picture.Lennox (1948) has pointed out that this large,

pale, eosinophilic cell should properly be namedafter Askanazy, who found it in toxic goitres inI896; and, in a most valuable paper, Lennox hasdescribed the incidence of acini composed of such

G2

April 1951 I9r

copyright. on 7 M


http://pmj.bm

j.com/

Postgrad M


pril 1951. Dow

nloaded from

http://pmj.bmj.com/


cells in various thyroid conditions. He foundthat these were:

(a) Rare in young subjects and in males of allages.

(b) Fairly common in otherwise normal thyroidsin elderly women.

(c) Present in 5 per cent. of non-toxic goitres.(d) Present in I6.33 per cent. of cases of Graves'

disease.(e) Present in 50 per cent. of cases of toxic

adenomata.(f) Present in all six cases of lymphadenoid

goitre studied.(g) Absent in four cases of Riedel's thyroiditis.(h) Present in the thyroids of myxoedematous

subjects only when lymphoid tissue waspresent, and absent when the gland waspredominantly a fibrous one.

It is of obvious interest that the Askanazy cellwas present more often in toxic than in non-toxicgoitres, and that it was constantly present inHashimoto's disease, but of particular importanceis the observation that it was almost alwaysaccompanied by and often imbedded in lymphoidtissue. This association was constant, except forsome otherwise normal thyroids which containedlymphoid tissue but no Askanazy cells, and foetaladenomas which contained Askanazy cells but nolymphoid tissue. Lennox suggested that theAskanazy cell-lymphoid tissue complex representeda phase of regression, through which hyperplasticacini passed before finally disappearing. Headduced some evidence to show that the cell wasof low endocrine function, and believed that, inhis case of toxic adenoma, spontaneous cure oftoxic symptoms occurred as a result of Askanazycell replacement or metaplasia of the original cells.

Friedman (I949) describes the Askanazy cell asa large, pale, eosinophilic cell, with a hyper-chromatic nucleus which is often bizarre in shape,and a well-defined cell membrane. He mentionsgranular and foamy types, and believes that he hasobserved its metaplasia from the normal cell type.He found it to have a distribution similar to thatdescribed by Lennox, and also noted its relation-ship to lymphoid tissue. In addition, however,Friedman found it in toxic goitres following boththiouracil and X-ray treatment, and in the thyroidremnants left after operation. In one most sig-nificant case, the progress of the histologicalpicture from hyperplasia to Hashimoto's diseasewas observed over a period of eight years.According to Friedman, the Askanazy cell is theresult of' cellular involution ' following prolongedhyperplasia.Summing up, it appears to be now possible to

state, with some degree of certitude, that:(a) Replacement of normal thyroid cells by the

Askanazy type is constantly associatedwith lymphoid infiltration:

(b) These changes are found most frequently intoxic goitres on the one hand, and inHashimoto's disease on the other.

(c) These changes are associated clinically withdiminution of thyroid function.

(d) The thyroid in Hashimoto's disease ishyperplastic in the sense that the totalamount of epithelial tissue is increased.

AetiologyIt is natural that many early writers, dealing

with thyroiditis as a single subject, should havesuggested an infective aetiology, and these viewshave been summarized by Lee (I935). Althoughno proof of infection was forthcoming, there weredescribed several cases of thyroiditis, includingthe Hashimoto type,-which appeared to followvarious infections and toxaemias, and Ewing (1922)considered that both Hashimoto's and Riedel'sdiseases were but phases of the same granulo-matous process. Graham's sharp separation ofthe two conditions discouraged further infectiveaetiologies for lymphadenoid goitre, although theycontinue to flourish as a widely accepted basis ofRiedel's disease.

Iodine medication was incriminated by severalauthors (Reist, I922; Boyden, Coller and Bugher,I935; and Dunhill, 1937), but there appears tobe little support now for this theory. De Courcy(1942) believed that perithyroiditis and consequentischaemia resulted in the fibrosis of Riedel'sdisease, but Goldberg and Davson (1948) alsoadduced ischaemia to account for the changes ofHashimoto's disease.

McCarrison (1928) was able to produce lymph-adenoid changes in the thyroid glands of rats fedon a diet deficient in vitamins; his findings havereceived scant consideration from clinicians, whosee no evidence of avitaminosis in their patients,but his experiments seem to be well worthy ofrepetition.Many authors have taken the lymphoid changes

of Hashimoto's disease to constitute the primarylesion; Hashimoto himself appears to have doneso, and considered that some relationship mightexist between struma lymphomatosa and Miku-licz's syndrome. Joll stated categorically that' theevidence is all in favour of the lymphoid changesbeing primary,' but had to admit that theyappeared to be 'neither inflammatory, neoplasticnor degenerative.'Graham (1931) thought that Hashimoto's

disease was the local expression of a constitutionaldisorder, and Crile (1948) has recently emphasizedthe evidence for this view. Schilling (I945)believed that epithelial degeneration was the

April 1951I92copyright.



j.bmj.com

/P


j.27.306.185 on 1 April 1951. D

ownloaded from

http://pmj.bmj.com/

REID: Hashimoto's Disease

primary lesion and that lymphoid infiltration wasa secondary replacement process, occurring in theburnt-out gland of an elderly woman. This viewdoes not take into account, however, what Jolldescribed as the intense intrinsic activity of thelymphoid tissue--an activity which seemed tohim to transcend any mere process of replacement.The existence of epithelial hyperplasia has

already been emphasized; Williamson and Pear-son (I929) ascribed the presence of many ' largeeosinophilic cells ' to proliferation of' subnormal'thyroid epithelium, while Parmley and Hellwig(I946) attempted to explain it on the basis of dis-ordered ovarian function, with not very convincingresults.

Other authors, including Eason (I928), Polowe(I934), Vaux (I938) and Goldberg and Davson(I948), have accepted a phase of hyperthyroidismas a significant aetiological factor, and have in themain conceived Hashimoto's disease to be theresult of exaggerated involutionary changes. Tothis view the histological observations of Lennox,Greene, Friedman, and Whitesell and Black havebrought a large measure of support, although aphase of hyperthyroidism cannot be regarded assufficiently common or sufficiently well establishedas to constitute a basis for all cases of Hashimoto'sdisease. Nevertheless, it seems now possible todefine Hashimoto's disease as a series of changeswhich include extensive Askanazy cell metaplasiaof the thyroid epithelium and lymphocytic in-filtration of the inter-acina tissues, which occur ina hyperplastic gland and which are associated withdiminishing endocrine function.Such is our conception of the pathological

mechanism. but before progress is made towardsthe synthesis of an aetiology, the followingquestions remain to be answered:

(a) What is the nature of the Askanazy cell ?(b) What is the nature of its relationship to

lymphoid tissue ?(c) Is a phase of epithelial hyperplasia a neces-

sary factor in the development of thedisease ?

(d) Can such hyperplasia exist in the absence ofclinical hyperthyroidism ?

(e) Would all thyroids which are the seat ofepithelial hyperplasia ultimately showAskanazy cell change and lymphoidinfiltration ?

(f) Can such changes arise in a previouslynormal thyroid?

Firm answers exist to few, if any, of thesequestions. The nature of the Askanazy cell, forexample, remains in doubt. Williamson andPearse (1929) and Parmley and Hellwig (I946)believed that the ' large eosinophilic cell' was. ayoung, proliferative and regenerative type, but

Lennox (I948) and Friedman (I949) describe itrespectively as regressive and involutional.Friedman has made a distinction between folli-

cular involution, comprising colloid storage inlarge acini lined by flat cuboidal epithelium, andcellular involution, consisting of Askanazy celldegeneration with its accoimpanying lymphoidchanges. He' believes that cellular involution isone possible end-result of prolonged hyperplasiaand states: ' Such morphologically demonstrablehyperplasia is not necessarily associated withhyperfunction in the sense of secretion.' Insupport of this statement, he cites the proliferativebut functionally ineffective goitre which resultsfrom either iodine deficiency or thiouracil medica-tion, and there is no reason to believe that thesenecessarily exhaust all the possible factors whichmight lead to hyperplasia without hyperfunction.

It is Friedman's view that Hashimoto's diseaseis the end-result of prolonged hyperplasia andcellular involution. His work suggests that thecondition may be a mode of thyroid failure, inwhich a tendency to diminishing function is atfirst off-set by epithelial hyerplasia, withoccasional slight evidences of hyperthyroidism,but, incapable of sustaining even a moderatelyhigh level of functional activity, the epitheliumpasses into a state of cellular involution. Furtherdiminution in endocrine function may then result,followed by the same vicious cycle of hyperplasia,involution and hypofunction in such acini as arestill normal, until the gland is diffusely involvedand the patient is clinically deficient in thyroidsecretion.

This hypothesis is largely speculative, but itpossesses certain merits; for example, it mightaccount on the one hand for the development oflymphadenoid changes in toxic goitre, and on theother hand for the development of Hashimoto'sdisease in a previously normal gland. It mightalso explain the existence, in Hashimoto's disease,of epithelial hyperplasia and endocrine hypofunc-tion, with, in some cases, fleeting evidences ofthyrotoxicosis in the early stages. There exist, ofcourse, certain serious objections to. this view.Hashimoto's disease is reputedly uncommon inthose goitre areas where prolonged epithelialhyperplasia is common and where involution isusually of the follicular type. Further, extensiveAskanazy cell change and lymphoid infiltrationhave not been described as resulting from pro-longed thiouracil therapy. Were it possible toassume in some individuals, as Williamson andPearse have done, a thyroid cell of intrinsicallylow potential, by reason of constitutional or extrin-sic factors as yet unknown, and that this cell wereincapable of responding in a functionally effectivemanner to even physiological demands, then the

April 1951 193copyright.



j.bmj.com

/P


j.27.306.185 on 1 April 1951. D

ownloaded from

http://pmj.bmj.com/


epithelial changes of Hashimoto's disease mightbe easier to understand.The mechanism of lymphoid infiltration remains

obscure. Jaffe (I937) believed that it was due toexcessive resorption of secretory products fromdisintegrating thyroid acini, but when Ferguson(1933) injected human colloid into the thyroidsof guinea-pigs, he found that the histologicalreaction was merely the non-specific one of fibrosiswith foreign-body giant cells. Although applicablein part to Hashimoto's disease, these findingswould hardly appear to account for all the lym-phoid changes involved. It is interesting to notein passing that Sheehan and Summers (I949)have described (after Simmonds) a' granulomatoushave described (after Simmonds) a' granulomatous'condition of the pituitary, which they liken to' pseudo-granulomatous nodules in (some) atrophy-ing thyroids.' The condition they describe consistsof endocrine failure, slight enlargement, epithelialatrophy and lymphocytic infiltration of thepituitary, with the appearance of giant cells,which they consider may repiesent groups ofdisintegrating epithelial cells.Shaw and Smith (1925) noted lymphocytic

infiltration of the adrenals in a case of Hashimoto'sdisease, and similar changes are present in theadrenals in 20 per cent. of those cases of Addison'sdisease which are due to primary cortical atrophy(Vines, I940). It is generally accepted that thy-roid and adrenal function are to some extent inter-dependent and, further, that there is a relationshipbetween adrenal activity and the behaviour oflymphoid tissue. It is conceivable that the pro-longed ill-health and 'hypometabolism,' whichCrile (1948) has emphasized in connection withHashimoto's disease, may in time be related tosome form of adrenal failure or dysfunction,although at the present time no evidence seems toexist for thia view.One further consideration remains, merely to

be intrbduced tentatively into this discussion. Itis reported (Rawson et al., 1941) that, apart fromthyroid epithelium, lymphoid tissue alone posses-ses any significant ability to inactivate the thyroid-stimulating hormone of the pituitary. It istempting to think that this observation may bearsome relationship to lymphadenoid change, butthere are at present no grounds for doing so.

TreatmentHashimoto's disease is not always clinically

.diagnosed as such and, in the absence of definitehypothyroidism, may be difficult to distinguishfrom nodular goitre and from carcinoma of thethyroid. The usual treatment, therefore, is sub-total thyroidectomy, and this has the merit ofremoving an unsightly swelling, eliminating or

preventing pressure symptoms and of affordingan opportunity for histological diagnosis. Post-operative hypothyroidism is usual and may becontrolled by thyroid medication.

Radiotherapy has been favourably mentionedby Renton, Charteris and Heggie (1938), Schilling(1945), Means (1948) and Crile (1948). McSwainand Moore (1943) suggested that hypothyroidismwas possibly less common after radiation thanafter operation, but there seems to be little evidencefor this view. Indeed, radiation has not beenemployed by Marshall et al. (1948), lest it shoulddamage any remaining functioning tissue. Therewould appear to be little advantage in radiotherapy,and it should certainly not be employed withoutbiopsy to exclude carcinoma.

In view of the existence of epithelial hyperplasiain Hashimoto's disease, it seems reasonable tosuggest treatment by thyroid alone; its effectmight include not only relief of the symptoms ofhypothyroidism, but conceivably also protectionof the epithelium against further hyperplasticchanges, and even its return to a normal restingstate. I can, however, find no record of a casetreated in this way. Thyroid therapy is, of course,almost invariably used after operation, and it isan interesting speculation that this may accountfor the rarity of post-operative recurrence of thegoitre.Hashimoto's and Riedel's DiseasesA strong body of literature has grown up

(Graham and McCullagh, 1931; McClintock andWright, '937; Joll, 1939; McSwain and Moor,1943; and Schilling, 1945) in support of Hashi-moto's original contention that the two con-ditions are separate entities, and this is the opiniongenerally held today (Aird, I950). The opposingview, that they are in some way related, has beenexpressed by Ewing (I922), Eisen (I934), Poer,Davidson and Bishop (1936), Vaux (1938) andMerrington (I948). Generally, these authorshave believed Riedel's to be a late stage of Hashi-moto's disease, and it was this theory in particularthat was attacked by Joll.

Earlier in this paper, under the headings ofincidence, clinical features and pathology, it hasbeen suggested that some of Joll's differentialpoints were perhaps of doubtful validity. Itremains true that Riedel's disease is more fre-quently unilateral, at least to begin with, andoccurs more often in males than does Hashimoto'sdisease. But the only fundamental distinguishingfeature seems to be the predominance of lymphoidtissue in the one condition and of fibrous tissuein the other, and the transition from cellularinfiltration to fibrosis is perhaps the commonestsequence of events in pathology. It is true, how-

April I951194copyright.



j.bmj.com

/P


j.27.306.185 on 1 April 1951. D

ownloaded from

http://pmj.bmj.com/


ever, that the transition from Hashimoto's diseaseto Riedel's has not been observed with certainty,although Heyd (1929) believed that he had doneso. Perman and Wahlgren (1927), Roulet (1931),McClintock and Wright (I937), Hellwig (1938)and Schilling (1945) have all obtained thyroidtissue from cases of Hashimoto's disease at inter-vals ranging from I8 months to 9 years, and in nocase found evidence of a histological change inthe direction of Riedel's disease.On the other hand, there have been reported

cases which seem to lie somewhere betweenHashimoto's and Riedel's diseases, possessingsome of the characteristics of each. Such inter-mediate types have been recorded by Schilling(1945), Means (1948), Merrington (1948) andOldfield (1948). It appears that there is, inMerrington's words, 'convincing evidence thatthe two conditions known as Hashimoto's andRiedel's thyroiditis can be present in the samegland at the same time.'

There may be merit in Womach's suggestion(i944)' that the two conditions are but varyingmanifestations of the same disease process, but

the problem remains unsolved, as it must remainuntil more facts are available. It seems likely,however, that the apparently decreasing incidenceof Riedel's disease will render its differentiationof less moment, and that future controversy willcentre rather on the aetiological aspects oflymphadenoid goitre.Summary

Hashimoto's disease is described and theliterature is reviewed. It is concluded that thecardinal features of the condition are epithelialhyperplasia, extensive Askanazy cell change andlymphoid infiltration, in association withdiminished thyroid function. The possiblesignificance of these findings is discussed.

AcknowledgmentsIt is a pleasure to record my indebtedness to

Dr. I. Doniach and Dr. Russell Fraser for theiradvice in the preparation of this paper, and toMr. K. G. Moreman, A.R.P.S., for thephotomicrographs.

rRBIBIOGRAPHY

ASKANAZY, M. (1898), Deutsch. Arch. f. Klin. Med., 61, 118.BOGART, cuoted by SCHILLING; discussion, LEHMAN, J. A.,

Third Int. Goitre Conference, Amer. Ass. Study Goitre, 1928.244.

BOYDEN, A. M., COLLER, F. A., and BUGHER, J, C. (1935),West. J. Surg., 43. 547.

CLUTE, H. M., ECKERSON, E. B., and WARREN, S. (1935),Arch. Surg., 31, 419.

COPE, O. (1948), MEANS, 'The Thyroid and its Diseases,'2nd Ed.

CRILE, G. (1948), Arch. Surg., 57, 443.CRILE, G. (1948), Ann. Surg., 127, 640.DAVISON, T. C., and LETTON, A. H. (1949), J. Clin. Endocrin.,

9, 980.DE COURCY, J. L. (1942), Surgery, 12, 754.DUNHILL, SIR THOMAS (1937), Trans. Med. Soc. Lond., 6o,

2462.EASON, J. (1928), Tr. Med.-Chir. Soc. Edin., 42, 169.EDEN, K. C., andTROTTER, W. R. (1942), Brit. . Surg., 29, 320.EISEN, D. (1934), Canad. Med. Ass. J., 21, 144.EWING, J. (1922), 'Neoplastic Diseases,' 2nd Ed., Philadelphia,

W. B. Saunders.FERGUSON, J. A. (1933), Arch. Path., 15, 244.FICARRA, B. J. (1946), Arch. Surg., 52, 729.FRIEDMAN, N. B. (1949), J. Clin. Endocr., 9, 874.GOLDBERG, H. M., and DAVSON, J. (1948), Brit.. . urg., 36,

41.GRAHAM, A. (x931), West. J. Surg., 39, 68i.GRAHAM, A., and McCULLAGH, E. P. (1931), Arch. Surg.,

22, 548.GRAHAM, J. (1939), quoted by JOLL.GREENE, R. (1950), J. Endocrin., 7, r.HASHIMOTO, H. (I912), Arch. f. Kin. Chir., 97, 219.HELLWIG, C. A. (1938), Arch. Path., 25, 838.HERTZ, J. (1943), 'On Goitre and Allied Diseases,' Oxford

University Press.HEYD, C. G. (1929), Surg. Clin. N. Amer., 9, 493.HORI, C. G. (1946), Canad. Med. Ass. .., 55, 372.JAFFE, R. H. (i937), JA.M.A., o08, 105.JOLL, C. A. (1939), Brit. J. Surg., 27, 351.KEYNES, G. (1939), quoted by JOLL.KOCHER, quoted by HERTZ, 1943.LEE, J. G. (I935), Arch. Surg., 31, 982.LEE, C. M., and McGRATH, E. J. (1937), Surgery, , 238.LENNOX B. (1948), .. Path. Bact., 50, 295.LUNDBAEK, K. (1942), Act. Med. Scand., 112, 55.

MARSHALL, S. F MEISSNER, W. A., and SMITH, D. C.(1948), New England Med. J., 238, 758.

McCARRISON, R. (I929), Bnt. Med. 7., I, 5.MEANS, J. H. (x948), 'The Thyroid and its Diseases,' 2nd Ed.,

Lippincott.McSWAIN, B., and MOORE, S. W. (1943), Surg., Gyn. and

Obstet. 76, 562.McCLINTOCK, J. C., and WRIGHT, A. W. (1937), Ann. Surg.,

106, Ii.MERRINGTON, W. R. (1948), Brit. 7. Surg., 35, 523.OLDFIELD, M. C. (948), Brit. . Surg., 35, 325.PATERSON, H., and STARKEY, G. (1948), Ann. Surg., 128, 756.PARMLEY, C. G., and HELLWIG, C. A. (1946), Arch. Surg.,

53, 190.PEDIGO, G. W., and ABRAMSON, W. (1947), Ann. Int. Med.,

27, 627.PERMAN, E., and WAHLGREN, F. (1927), Act. Chir. Scand.POER, D..H., DAVISON, T. C., and BISHOP, E. C. (1936),

Am. .. Surg., 32, 172.POLOWE, D. (I934), Arch. Surg., 29, 768.RAWSON, R. W., STERNE, G. D., and AUB, J. C. (x94I), Tr.

Am. Ass. Study Goitre, p. 159.REIST A. (1922), Zeitsch.f. ath., 28, 141.RENT6N, J. M., CHARTERIS, A. A., and HEGGIE, J. F.

(1938), Brit. J. Surg., 26, 54.RIEDEL, B. M. C. L. (x896), Verhand. d. Deutsch. Gesellsch. f.

Cdr., 25, 101.ROULET, R. (I931), Virch. Arch. 280, 640.SCHILLING, G. A. (945), Surg., Gyn. and Obstet., 81, 533.SHAW, A. F. B., and SMITH, R. P. (1925), Brit..7. Surg., 13, 93.SHEEHAN, H. L., and SUMMERS, V. K. (I949), Quart... Med.,

8V

SIMMONDS, M. (1923), Virch. Arch., 246, 40.SIMMONDS, M.(3i rCh.Arch.,a46,4o.SINAIKO, E. S LAWRENCE, C. H., and ZITMAN, I. (1948). Int. Col. Surg., , 62.

TAYLOR, SELWYN (I9x), Personal communication.VAUX, D. M. (I938), .B. Path. Bact., 46, 44.VINES, H. W. C.,' Green's Manual of Pathology,' i6th Ed., p. 832,

Bailliere, Tindall & Cox, London.WEGELIN, quoted by HERTZ, 1943.WILLIAMSON, G. S., and PEARSE, I. H. (1929), Brit. Med. J.,WHITESELL, F. B., and BLACK, B. M. (1949), J. Clin. Endocr.,

9, 1202.WOMACH, N. A. (1944.), Surgery, I6, 770.

copyright. on 7 M


http://pmj.bm

j.com/

Postgrad M


pril 1951. Dow

nloaded from

http://pmj.bmj.com/

Documents

HASHIMOTO'S DISEASE - pmj.bmj.com · i85 HASHIMOTO'S DISEASE: A REVIEW By DOUGLAS R. K. REID, M.D., F.R.C.S.* ' If wesummarise the present clinical findings, wehadfourpatients, all