sarcoma of breast, with particular reference to its origin ... · fibroadenoma with cellular...

J. clin. Path. (1962), 15, 1

Sarcoma of breast, with particular referenceto its origin from fibroadenoma

R. C. CURRAN AND 0. G. DODGE'

From the Depar-tments of Pathology, St. Thomas's Hospital Medical School,and Sheffield University

SYNOPSIS Thirty-nine sarcomas of breast are described. The patients' ages ranged from 26 to 78years but most patients were middle-aged or elderly. Evidence is adduced that 16 of the tumoursarose from pre-existing fibroadenomas and others very probably did so. Many tumours wereadenosarcomas and one was a carcinosarcoma which preserved its structure in a lymph-nodemetastasis. Recurrences were mostly local, but a small group had a much graver prognosis. Apleomorphic and giant-cell structure usually meant a high degree of malignancy but the significanceof cartilaginous or bony metaplasia, present in seven tumours, was less certain.

It is wise to regard all 'fibroadenomas' of the breast occurring in patients over 40 years of age asat least potentially malignant, and these shouLld be examined histologically with care, particularlysince sarcomatous change may be present in one part only and therefore easily overlooked.

This paper reviews those cases diagnosed as sarcomaof the breast in three large teaching hospitals duringthe past 25 years. Although this type of tumour hasproved to be comparatively rare, in that only 39cases were found, the most striking fact to emergewas the close relationship which often existedbetween the sarcoma and a pre-existing fibro-adenoma. The presence or absence of this relation-ship provided in fact a useful way of classifying thetumours into (1) malignant fibroadenomas, in whicha new growth of sarcomatous, adenosarcomatous, oreven carcinosarcomatous tissue had developed froma fibroadenoma; (2) sarcomas with included epithelialstructures (adenosarcomas), in which a fibroadeno-matous origin was less certain; and (3) 'pure'sarcomas, which had no detectable link with pre-existing fibroadenomas.Two features of this scheme of classification should

be made clear. First it is exclusively histological,since clinical assessment of the nature andpotentialities of breast tumours consisting largely ofconnective tissue is rather inexact, probably becausesurgeons rarely encounter them; and secondly, thereis no hard and fast division between the groups andthe distinction between groups I and 2 is particularlytenuous.Received for publication 3 July 1961

'Present address: Department of Pathology, Makerere College, P.O.Box 2072, Kampala, Uganda

RESULTS

The main features of the 39 cases are listed in TablesI, II, and III.

MALIGNANT FIBROADENOMAS

The malignant fibroadenomas (Table I) are 16 innumber. Basically, these are fibroadenomas with astroma which has become sarcomatous; 10 of thefibroadenomas are mainly intracanalicular ('giant'fibroadenomas) (Figs. I and 2) and six mainlypericanalicular in structure. Seven of the cases areboth histologically and clinically malignant; sixare histologically malignant; and inz three cases(Nos. 2, 3, and 4) the evidence of malignancy islargely clinical. In some of the 13 tumours showinghistological malignancy, the sarcomatous part ofthe tumour is relatively small and it could have beenmissed if only one or two sections from each tumourhad been examined. Of the 16 patients, four hadrecurrences in the operation site; one of these fourlater died of sarcomatosis, and so did two otherwomen. Only one person developed secondaries inthe axillary lymph nodes and her tumour was,significantly, a mixture of adenocarcinomatous andsarcomatous elements (Case 15). Although most ofthe new growth is sarcomatous, the epithelium toois partaking in the process in several tumours and

on May 17, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.15.1.1 on 1 January 1962. Dow

nloaded from



j.com/

J Clin P


nloaded from



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

http://jcp.bmj.com/

http://jcp.bmj.com/

R. C. Curran and 0. G. Dodge

TABLE I


No. Age of Length ofPatient History(yr.)

RelationtoMenopause

-Site MacroscopicAppearance anidSize

Histology

1 78 3 mth.

2 69 40 yr.;growing for6 mth.

3 65 5 mth.

4 62

5 62

6 61 6 wk.

7 59 8 mth.

8 57 6 wk.

9 57 6 wk.

Alter Upper halfright

5 cm. diam.

After Behind right Mucoid and cysticnipple mass 5 x4 x4 cm.

After Upper leftouterquadrant

After Above leftnipple

Hard tumour fixed toskin, 6 cm. diameter

Ovoid mass 6 x 3 5 cm.

After

After Upper left Irregular shape.outer white and fleshyquadrant with cyst formation

6 cm. diameterAfter

After Mid-line,above rightnipple

After Right

Smooth tumourattached to skin andmuscle, 8-5 cm.diameterFreely mobile firmnodular mass, 2 5 cm.diameterGelatinous tumourwith small cysticspaces 3 cm. diameter

Hyalinized pericanalicularfibroadenoma with partsconsisting of ductsembedded in cellularspindle-cell stroma ofsarcomatous type? Myxosarcomatous changein an intracanalicularCgiant') fibroadenoma

Intracanalicular ('giant')fibroadenoma with cellularstroma, not definitelymalignant; calcificationand bone formation instroma

Intracanalicular ('giant')fibroadenoma; a fewareas of cellular stromaMalignancy moderate ordoubtful

Myxofibrosarcomatouschange in an intra-canalicular ('giant')fibroadenomaSarcomatous change infibroadenoma

Myxosarcomatous changein an intracanalicular('giant') fibroadenoma

Sarcomatous change inan intracanalicular ('giant')fibroadenoma

Myxosarcomatous changein an intracanalicular(giant') fibroadenomaOsteoid present

Simple mastectomyWell 6j years later

Local excision. Localrecurrence ofcystic andgelatinous sarcomatoustumour, 2 years latertreated by simplemastectomyRadical mastectomyDied 5 years laterfrom sarcomatosis

Simple mastectomyDeath from sarco-matosis 4 years later

Radical mastectomyWell 3 years later

Local excisionWell 1.1 years later

Simple mastectomy,radiotherapyWell 6 years later

Local excision

Local mastectomy,rad iotherapyAlive 3 years later

forming abnormal glands: these tumours are there-fore adenosarcomas.

ADENOFIBROSARCOMA

Scattered throughout the sarcomatous tissue ofadenofibrosarcomas (Table IL) are numerous glandu-lar structures. In a few tumours these are scanty(Cases 22 and 24) and it might be postulated that theyare pre-existing ducts 'caught up' in the sarcomatoustissue. But in most of the tumours their closeintegration with the sarcomatous element suggeststhat the tumours are true adenosarcomas. There isno benign fibroadenomatous tissue, however, andtherefore a convincing fibroadenomatous origin islacking. All the same, the readiest explanation is

that these neoplasms are pericanalicular fibroadeno-mas which have become malignant.Of the 12 patients, three had local recurrences and

another developed a solitary lung metastasis fourand a half years after resection of the primary.

'PURE' SARCOMA

The 'pure' sarcomas (Table HII), 11 in number, aremalignant mesenchymal tumours in which noepithelial structures can be found. In two patientsthere was local recurrence which was successfullytreated. Three died of sarcomatosis and two othersmay have died from the same cause. One is stillalive but has multiple secondaries.

Treatment andPost-operative Couirse

2



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

Sarcoma of breast, with particular reference to its origin from fibroadenoma

TABLE 1-cont.


RelationtoMenopause

Site MacroscopicAppearance andSize

Histology Treatment a'idPost-operative Course

10 56 10 wk.

l1 50 2 yr.

12 45 5 mth.

13 43 6 mth.

14 40 1 yr.

15 38 2 yr.

16 35

After Upper right Well encapsulatedouter mucoid, 6 cm.quadrant diameter

After

? Upper leftouterquadrant

Large, irregular,somewhat translucentand gelatinous mass(10 x 14 cm.) showingextensive necrosis andsome areas ofhaemorrhage

Over 41 cm. diameter

Before Left Lobulated gelatinousmass filling breast

Before Right Encapsulated 8 cm.diameter

Before Left Lobulated encapsu-lated mass withmucoid and cysticareas13 cm. diameter(700 g.)

Before Left Rubbery, partly,encapsulated greyishwhite mass6x4x2 cm.

Sarcomatous change inan intracanalicularfibroadenoma

Sarcomatous change inan intracanalicular ('giant')fibroadenoma

Sarcomatous change infibroadenoma

Intracanalicular ('giant')fibroadenoma with cellularstroma of moderatemalignancy



Sarcomatous change inan intracanalicular('giant') fibroadencna

Simple mastectomyLocal recurrence of'adenosarcoma' 5years later, easilyresected, very similarto primary in histology

Local mastectomy

Local excision, thenradical mastectomyWell I year later

Local excision oftumourMassive local recur-rence of adenosarcomaafter 17 monthsRecurrence excisedWell 2 years after firstadmission

Local excision andradiotherapyWell 24 years later

Simple mastectomyRecurrence in scar(sarcoma) and axilla(carcinoma andsarcoma) after 9monthsDied 2 years after firstoperation with axillaryand lung secondaries

Local excisionPost-operative radic-therapy

ILLUSTRATIVE CASES


WITH LOCAL RECURRENCE A woman of 43 (Case 13,Table 1) presented with a lump in the left breast whichhad been growing for 6 months. She was married but hadno children. Menstruation was normal. A firm mass

almost filled the breast but it was not attached deeply or

to the skin. Local excision was performed. The excisedtumour measured 7 x 6 x 5 cm. with outlying satellites upto 05 cm. diameter. Parts of the tumour had the structureof a 'giant' intracanalicular fibroadenoma (Fig. 3), but inspite of histological evidence of malignancy (Fig. 4), thelesion was diagnosed as benign. However, the patientreturned 17 months later with a massive recurrence at theoperation site. Simple mastectomy was performed. The

tumour's structure still resembled that of the primary butthe connective tissue component now showed a muchmore sinister histological pattern which was franklysarcomatous (Fig. 5). Numerous glandular structuresmingled with the sarcomatous tissue, and, interestingly,occasional mitotic figures were present among theepithelial cells (Fig. 5). The patient was well and free fromrecurrence six months after the second operation.

This case probably illustrates the reluctance ofpathologists to label a 'fibroadenoma' malignant.The tumour was very large and there was histolo-gical evidence of mitotic activity and of localinvasion, but the possibility of malignancy was notconsidered. However, there was fairly rapid localrecurrence of adenosarcoma.

3



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

wo~~~~~A.. .

a' * - A _.,..r.b

4..~.

Va'.,.N'1,.

:, ..* S,

*r eo

> ~~ v4>~9 A

FIG. 1. Case 8, Table 1. The large intracanalicular process, typical of the true 'giant' fibroadenoma, consists of sarco-matous tissue covered by a thick layer of hyperplastic epithelial cells. The cystic space which the polypoid mass fills islined by a very attenuated epithelium. Note the compressed normal breast at the top. (Haematoxylin and eosin x 60.)

FIG. 2. Case 8, Table L.This is a sarcomatous area.The cells show great pleo-morphism; most are spindlecells but some are largeround pale cells (bottomright). There are two mitosesin this field (top centre andbottom centre). (Hae-matoxylin and eosin x800.)

EL :'ar.,.! r

_ .....

._.*_L

,.

:s.g^' B::ii ........ ,'.:;:.:.. ....

.: t.e :.

... :-

_ i'9' wFE

...:;..S:'.z .s

....

.1

*w.. Xx

!

i...41 ll.

- :.:. I

.ip

";.: "AW



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

Sarcoma of breast, with particular reference to its origin from fibroadenonma

X, ',''''''b-il

sEr,jSs[;Ket,svt S 1Er'A,V's XtXs

FIG. 3. Case 13, Table I. Numerous irregular glands are scattered throughout the sarcomatous tissue and the patternof this part of the adenosarcoma is therefore 'pericanalicular'. (Haematoxylin and eosin x 100.)

WITH BOTH MESENCHYMAL AND EPITHELIAL ELEMENTS IN A

LYMPH NODE METASTASIS A woman of 38 (Case 15,Table I) stated that she had had a lump in the left breastfor two years. She had one child, age 7; lactation hadbeen normal. On examination the breast was filled by afreely mobile cauliflower-like mass and axillary lymphnodes were enlarged. Simple mastectomy was performed.The tumour, 13 cm. in diameter, weighed 700 g. It waslobulated and encapsulated, but there were areas ofmucoid and cystic degeneration and of haemorrhage.The more solid areas were soft and pink. Histologyshowed that the tumour, though basically an intracanali-cular fibroadenoma, consisted in places of a complexmixture of fibromyxosarcomatous tissue and abnormalepithelial structures; these consisted of small acini andislands of atypical epithelial cells, sometimes of a squa-mous cell type (Fig. 6). The patient was readmitted ninemonths later with a lump the size of a marble at theupperend of the mastectomy scar; there were also two massesof enlarged axillary lymph nodes. These lesions wereresected. Histologically, the nodule in the scar had thesame structure as the carcinosarcomatous part of the2

original tumour. The lymph nodes were more or lesscompletely replaced by tumour; in some, this was aremarkably well-differentiated spindle-cell sarcoma (Fig.7) but in one the deposit was a mixture of sarcomatoustissue and malignant epithelial acini (Fig. 8). Despitepost-operative radiotherapy, there were further recur-rences in the scar and a fungating mass had developedin the axilla 14 months later. A large rounded secondarydeposit was also present in the right lung. She died amonth later, two years after the first operation. Therewas no necropsy.

This tumour was unusual for two main reasons.First, it seemed to be still encapsulated at the firstoperation, even though it was unusually large.Second, the histological structure of this primarygrowth was a mixture of sarcoma and carcinoma;and even more surprisingly, this structure was laterperpetuated in the secondaries in the axillary nodes.The carcinomatous element may have been thereason for the lymphatic spread, since in none of the

5



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


A:..

i.:*: -1-.

"4

It', ;. 4

A

*

t I

FIG. 4. Case 13, Table 1. Well-differentiated fibrosarcomatous tissue (left) is separated by a dense fibrous capsulefrom acompressed normal breast lobule (right). The capsule was present round only part of the tumour. (Haematoxvlin and eosinx 120.)

:.,

*.....m ' :

*

FIG. 5. Case 13, Table L. Adenosarcomatous recurrence of tumour shown in Fig. 4. This is a wel -diffrrentiated andfibrous part of the sarcomatous tissue. An elongated glandular space runs along the top; its thick epithelial lining containsone cell in prophase (top centre). (Haematoxylin and eosin x 600.)

6

:. :1:.

'..' ;W

...j...

?§-... ''.4, .1SoCt-:

iNikw, VWWOW

410

Agto

*_ -..*a -

...Altisi."ll ...7.11.,.MM

'41IIIIIIIII&W



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

FIG. 6. Case 15, Table L Tke sarcomatous tissue is well-differentiated and fibrous. The small clumps of epithelial cellsshow squamous metaplasia. (Haematoxylin and eosin x 160.)

FIG. 7. Case 15, Table L Sarcomatous recurrence in an axillary lymph node of the tumour shown in Fig. 6. On the leftis the main mass of well-differentiated and remarkably fibrous sarcomatous tissue. Penetrating the lymphoid tissue (right)ahead of it, and completely surrounded by lymphocytes, is a smaller deposit ofsarcoma. (See also Fig. 8.) (Haematoxylinand eosin x 85.)

4 .. I

i.:.."Ia,1.

j.,:0



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


FIG. 8. Case 15, Table I. Carcino-sarcomatous recurrence in an axillary lymph node of the tumour shown in Fig. 6.Carcinomatous glands are penetrating the lymphoid tissue (left) ahead ofthe main tumour mass (right) which is a mixtureoffibrosarcoma and adenocarcinoma. (Haematoxylin and eosin x 90.)

other 38 cases were lymph nodes found to beinvolved.

ADENOFIBROSARCOMAS

WITH SOLITARY LUNG METASTASIS FOUR AND A HALF YEARSLATER A woman of 64 (Case 20, Table II) presentedwith a lump in the right breast which had been growingfor five months. On examination it was 'the size of anorange' and situated in the upper outer quadrant. Therewere attachments to the skin and deep structures, andthe overlying skin was inflamed. A radical mastectomywas performed. The tumour was a firm, apparentlyencapsulated mass, 6 cm. diameter. Microscopy showedthat it was a fibrosarcoma in which were embedded manyglands (Fig. 9). The sarcomatous tissue surrounding someof these ducts was remarkably basophilic (Fig. 10) andsometimes resembled hyaline cartilage (Fig. 11). Necrosiswas extensive. The axillary nodes were free of tumour.The patient remained well for four years, but four and ahalf years after the operation she returned to hospitalwith a history of loss of weight of 14 lb. in the previoussix months. The mastectomy scar and the left breastwere normal but a chest radiograph showed a singlerounded shadow in the right lung for which pneumo-nectomy was performed. The excised lung contained asolid tumour which replaced the middle lobe. Histolo-gically this was a well-differentiated spindle-cell sarcomalacking epithelial elements. The patient died four days

after operation. At necropsy there were no othei depositsand death had been caused by multiple thrombi in thepulmonary arteries.

On clinical grounds this patient was thought tohave a carcinoma of the breast and therefore aradical mastectomy was done. This may be whythere was no recurrence in the mammary area andperhaps also why metastatic recurrence was delayedfor four and a half years. On the other hand, the factthat metastasis took the form of a solitary, well-defined mass which was easily resected suggests thatthe tumour was not highly invasive and thereforeless liable to produce multiple secondaries.

WITH TWO LOCAL RECURRENCES A woman of 58 (Case22, Table JI) reported with a lump in the right breastwhich had been present for one month and which,though hard, was mobile. Simple mastectomy was per-formed. Microscopy of the excised tumour (6 cm. dia-meter) showed a spindle-cell sarcoma with fairly scantyepithelial elements. The patient remained well for somesix months but within nine months of the operation amass 5-5 cm. diameter had developed at the operationsite. This was excised and microscopy revealed a some-what higher degree of malignancy. Two months afterthis episode she developed another recurrence in thesubcutaneous tissue above the operation scars, and

8



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

Sarcoma of breast, with particular reference to its origin fromn fibroadenonia


RelationtoMenopause

TABLE II

ADENOFIBROSARCOMAS


Histology

17 78 10 mth.slow, thenrapid growth

18 67 2 mth.

19 67 4 mth.

20 64 5 mth.

1 61 10 mth.

22 58 1 mth.

23 54 3 mth.

24 50

251 39 21 mth.: in-active for 20mth. then rapidgrowth for I mth1

26 34 6 mth.

After Lower leftinnerquadrant

After Left

7 5 cm. diameter

4 cm. diameter

After Outer right Tense and cystic,12 cm. diameter

After Right

After Upper leftouter quadrant

After Right

After Right

After

Before

Before(lactating)

Right

Right

27

28

Firm, encapsulated

mass 6 cm. diameterfixed to skin andmuscle

Round, well-circumscribed andfleshy, 6 cm. diameter

Firm, round, circum-scribed, 2-5 cm.

diameter

Mobile tumour12-5 cm. diameter

Yellow, rubbery,spherical tumour,8 cm. diameter,encapsulatedSolid, encapsulated,5 cm. diameter

Whole breast en-larged (20 x 18 x 12cm.) and replaced byfleshy and cystic massA small firm tumour(1 cm. diameter) incentre of mass

Spindle-cell and giantcell sarcoma

Sarcoma with giant cellsand osteoid and boneformation; squamous meta-plasia in epitheliumFibrosarcoma

Low-grade fibrosarcoma,extensive necrosis, manyducts

Spindle-cell sarcoma

Well-differentiated spindle-cell sarcoma with only afew ducts

Fibrosarcoma

Osteochondrosarcoma withducts around edgeSpindle-cell sarcoma withgiant cells and chondroidareas

Spindle-ell sarcoma withgiant cells in lactatingbreast

Low-grade fibrosarcoma;ducts around peripherySpindle-cell sarcoma,small abscesses

Radical mastectomyLocal recurrence 3months laterRadical mastectomy

Simple mastectomyLocal recurrence anddeath 6 months laterRadical mastectomySolitary sarcomatousmetastasis in rightlung resected 4j years

laterDeath from pulmonaryartery thrombosis(post-operative)Radical mastectomy

Simple mastectomyLocal recurrencesafter 9 and 11 monthsRadical mastectomyafter 13 months thenradiotherapyNo metastasesWell 2 years after firstoperationRadical mastectomyWell 2j years later

Radical mastectomy

Local resection

Simple mastectomy

Radical mastectomy

'On waiting list for 20 months to have 'fibroadenoma' removed before sudden growth necessitated admission

finally, 13 months after the first operation a radical well, with no evidence of local recurrence or of metastasis.mastectomy was performed. The excised breast containeda firm tumour, 3 cm. diameter, which had invaded the Histologically this was a sarcoma of moderatepectoral muscles and histologically was a spindle-cell mlgnancand the was beharcoma of tesarcoma. There was no evidence of chest secondaries, malignancy, and the subsequent behaviour of the

Radiotherapy was given to the mammary area post- tumour was in keeping with this interpretation;

operatively. Ten months after the radical mastectomy there have been two local recurrences but, so far,(23 months after the first operation) the patient was no distant metastasis.

Treatment andPost-operative Course

9



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

... .0 .4. m- .:::p ..-...:.....:..

N; 1. .:.. ..i.8i?.- .-...11-1*.........,., .....f..:. ......,6..-*w-..'t..-:.-,,"-......-RA..*--+......I-.:.,1,-.-9.'......If.---.I-..,,.-;I.--:--:.II!,.. ,..f.;....,.. .: kf,,I.-..-kb.-..I.......;!,I::!4........,". .,.:.-+. .,.I...--,--.*: II.-:I-:.: .:,. 1-:.:.."I"I" #:,:.4 :-*-I..,11 .:!.-, !.*......-".:.... ng. ......:..--4..-.....;i..:.....-4r.,.... :;. I -...":

-;:I.:.?.-!' j:X.::...:.... ,:.::1-:,", -.... C,'f", :jl. --ft.0...0..-,F ..:::...,....1,44 .1 --W ,:O, 1. ..,:.....-A::--,.-.:'"'..::; .,". .......;-................. 7.1..1-i......:.. i. ".

.... .... ::- ..: :.E. -90'k

-.1;;

.::.:::::.,i,F. ........., -..,..,. -Ir--,.,,:.;!Ir .?-0:-,.I......... -.'.

;.. :.. ... ......:.-.- ..,:. .....A..--:........, :..:... ...:,

:."- ".... .I.-.-:,-,t.." I.:.: ;:--- .:,f....,

,:: 'O-:. ..e....--.-A.: :.*.III.i, J....I ---Lj.-... ...-:-:-.

.:.:.., -:; .....O!",,-...:;:;. -...;. -:1 ..:!." 1.'...-! -1: *-I" :!.: .A---.....,-.0 4-i..;:....:.:.;,x!::...- ....7...;..:::..:...:-.,... .1.F.i...:0.q.:::.--;.... +, --I-*V..g;.?.:.-:;...:.....-x.. .1..-%-.......:: ....:F, "i;-........... .1 :.I.-P...-C. ".k.,..-..-... .!!. ;.0-::..;.:::...- hf.... ....:.: A.... .W....::.,:;W.p ...-.....'...--0.-,ft !...... -.I....-*-.R: .:".:!I,...-.:.,".. .,.......O:.:. -:-.-'-'-'---..I:0 ,.:O:.,...'a 4. .., 11..I:

....,......-V.N..%"K. .:I.Ijl..:- .....hI.,.lxm..4 .: ......:......j. ..... ..

:............:.. -1%.:#..-4-.-..-.-:.:: :.::......A-,ti.-,.......

..I.: -.--.:t-,::.........O":.1-P..... .:... t.. T.A: #..:!:::.: ....:,R. ,-..-'.'. ... ..:.i....1..::. J85 ,". 'e, ':X. -f f.. ..r-.-.: ..,`.'...-.*..: 'i-...iF.... -4.i; -..; 12.,.....t..-:1 'C............IE....j..k'0...-......:,_.:: i".I.... At'. :%:i:. ,4. I.- :%:.-.:..-:I'::..:%. ..-,j, .F .1 ..... ,,.-iv....-.:P .:.,...:::...: ,.:: ..! ".4 O..:&... ... :1 ........I..V-.-1 ...,,,I,k ::X.:"..i..,.--,:: ':'.,:: ]&*, "..: ....9...".....-- ..: ...-4... -!.-A1:U.:.. ..-:.: .3... .. :x;..:i:I.:.;.-0-:.;.-:.''';..:::.;;..::.;...:.--...-,:...,.:.::f:..... -. ..,,..:.....:......-:-I:-I....it :-.:... -:,.: :-,.A...---:..I:...:. :1

;; ..:.-....:.I-..:.:: ...-.-..0,-.I....,,-/:% .,&- ;.::. .-i,4 ..-.....:.-.:: ",,,.4 :§:.......::.'........j....-."' .:.:.fZ...:s:::. .;F.: 0i*... ........-.:4fI4v----:- ..:.::AI.:;!.:,..... .C........,.. I.....::, t. ...-:,.-..........,., .:. ,, .7. ,;,.:k.....e. .PI.;,; ..:...:::--.,.,-""C,"...." '.-4V..:, -K ..,

..:q.-.1.--"4.>. ....::.-..:.,,:. -IN .:!..1..... j. -,." s.........: -M .% S.-.-.--f4, -Y -i"1, 54","'::.: ,A:.,:` :.` ......'01.,..J:--...:. ..:;:,:....',,,,

`.--.-i: -:....A5i: :::.,0.. V',.."'.I,I",.,I.: -o _.-.-.i*....j.,,,.. ....--.. :-. -.%'-,,,.......0"' "'......0.!.. ,.!...I..41 -#:- :.I4":' .e` ,"":::,.....-..::I.f -j .. ... ....,-,:.X.F;...x .§..:., .t.r.,.,..:%-..I.....:...11.1 ". - -.; ..:::..... .V .:..i-.%.1' .zp..-,...-.,. ..' 'f,...",,,:: .`%.II:,.i!. .-.:.... -X 1::1U. -:. ..:1.-..:- ., ..,P.-.:

..... ;( .A., -I.M g::-.*-.*- f.. -'-' '* ./ -:-.......N.t.-.... ....-'...,I...,...',",'%#----4-.:-...,-, ..-."....;:-1:I-, .., -9 .1. ."' -r .:.;...,., ,I", ..-.."-..........X&.;'-..I.--,A......I..: ,,:--:....O' -I .11, ': .;R,-;.I:,?.. .- .-:- 1-:.::-K-. kj'......:.4, .-..- ..,--"- "' --. :.- %... o ., ";, --I-- t..

.4.,"'."'...`..,f","...,Ii i.:f...I..:1 ....-.;I lw ;..;.,.....--..:.I"......-v' ...-....-.'....:,-..4-..-.I----.,--?, -..----...--..-,-. A-M,..--il-- ----........;.. ..I---.,-..111 ...i .;: :.I.f..... -F:--II11.1.: ---,--:'t -..:I."::'-.*.i..... It

..?,

....,"...-...,-i--..::..:t:...---.aII ..;:. O

1.-:- .-I .;,.-:::.; ....,4...;-:."'-,::.: ,-..-..'.!.." I --, -. ..:....I,......i:..,,... :..II. ....rti, .: -, ie .."

.1..:. ...--.....-.A-........,..,::, .--' --.:.`,.: ... ......r.- ..-I. :.A... :.....f....:4........: OF.:-......*..--.:.".;...,i.. ..-W .z:.:....j:...,.... .!.::. .'. ------- X....,..;. .!.! +..,i .'.'.:-... ... :..,..!!:..P.....!:#....:.-;: ....:F-tf-- -,-.-..: ...:,,.::.,..I", :-, .-,.1.".. ..'. 1. -1:1: -:1::..,....,.. .!k.%.:...Y1..I' ::.

...1.:.. P4 ::,-: :;.. -:.-,,ol: '% '.: 'ile,.........

O .,A,,::..-.z..-:-.!...... :.,,.: -,;..W...:..:...::: i. .:: ....:. si :.' '..".", :-.- ..,. -.0-:::.::-,..... *-.-%..:!..19-:..",-.'.'.",.:.!:,.!:P-:-... ..!.i4- f. --!

'...:O...*..II.oI.... -k

,.A-'.... ...-- -:.::. ... .....--.--*4...I...:;..ii-. 4:- ..'....hO...4:?...1,:.........,..0.1154.-,.."-:- :k*.-IA 6:..... %; .Cv:.::::;I:- .:i-...-..., ,F.;. a.,.,,:.::..:..5!.. .!.:,'..i.::.:..:-:...I,,..--f;..-.:.-,.:O;.:. #.,--.%...%,.i:!:,.-.--,....:.".: -:- .. ::: "".A.tIf ::F,. .6

.:.:I .:..::. ......P ;.. .!.I'll .0- ;W..--;..X......1..%;..

......0..4. ........

%-F."..V....-:-,-,.: .,;:.....---,... -,..........*..# A..,....I-:..-..%. ..t'',-I-:I...,,-.,..... MO,i.-,---'. ir,......'... .S.4... .::;ii.1 .., -.1 ,...,...% 9,11N-.,:.o ....:V..:. :,:.: 1:....9.?. ....,..:. :, .4. ..,-.*.f;..-....-i.-.,I

.'A r- ..T...,::4.. -:--..-..:...:.. ...:-I.:.W.W.-::. -f.-:.-, t--:.411.... :8 :.,:.-'-'-" :..,.... .:. -F.

....I1:%...... ... '-I...... *O'..:,.4-*....."',,....;: -I-"...I- ".. ....%I- .;. .. *.j.iI.... :, ---.:.....,:;,*",:.::....I,

..#..".:t- T..... ki-.?. .I..V..k----:-.: -.:.. J:::.-.I.....lo.......i.I-.........--J-,- -... -:.:..I...i...#v...-:.;.,......1%.IIf A.,-:-.-':--.,-4:- ,:,, :...1-...X::: ......I..J:...-.-...... N ...... ...V.. .....:.. .-" "'..tl'.:4-k.....:"" i. $::: ..::

.. 1....-,...:I...1. ,.. ..A----; ..:--.:....: ':-.,..4. ...'! .-I-,-I-v.v....A:-,I. -,:-#, . ,40 %:,,.:,'.-.:.:...1:..

..,: ...;....: I.- .4...L C. '. ..-:,:.,".....-.::..:::.: :g.I.: ::-..A.I:.:.....:.....:--.-.-.%....:%" .;:..:. .:.#- .:....:O: .::....,.,:..,:-....0 ..r-:......:...:1:i!

...::.#" .PI". .: :.:,.;:.."-1....'-!- -i.....iI.,;:.!" 'O .. .:-,, :::k.:: ;:.:,:f.1 ..-:.I... ...'A. ffl, if.:R:... !:--'.-..:.....".: t. -:.-..4I..,,...-II I!:jO -:: -:..... .k. .....:.. ::...:.--,,,,.......-..; .:o.;I.-.. W -F.- .:.;....I..46k....:,.:.APkM..i,..,. ;;....-I .:N..' -.14A......,,.--,.,.......-.1 A. ..::O: A........:;:,:;%:!K'' ........,:44 ....::.I...

.......:..:.! ..;F..*.-...,-.:. 1,14 ..--;--:,.:.::.T. II .......#'... .:-, ....I.. ::.:..::.: O... .: I ......:.......: 1,"".,..,.":...-e.."..""....;O 0-1..i. ..-,-.. :..e.:..4%:.::.... .. ...:,..!.:%-.-i::..t:...,:...,.4::, 41-1. 41 x,:--""-,,-I.r,..,- -..R..........::...,.,..1i.

.-.- ;.! ....:8...:,,,,, ,...f."4-.F ...!;...I,-""'"!,t: "..:;,.......,,...;t....k......i.I#.-. ...-RK.,. --..I C,..;.!: .i:::,. ..A:., ..E.

..... :!.,,"",J,.. .....-.-,:::..-...1...I:..W.., .............---..o......%A. ..."A'... '.... ..- '3,1.. ,. *-..,...,,4:..., -;* --- e!.-.--.444.-U-,,".f.,:..-"!",-.i. .''I .-":".,: r: .!:, -0:.9-;I &%"% ftf....0-...#..-*-:-*%:-,..-'L.-,.-w--:- r,,7 ;u:

.*:::.;- A, :.,I,

FIG. 10. Case 20, Table II. A group of normal breast ductules (left) are being compressed by the growing edge of the

adenosarcoma. The sarcomatous tissue is more cellular and basophilic around the glandular elements of the tumour

(centre). (See also Fig. II.) (Haematoxylin and eosin x 90.)

1-



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

Sarcoma of breast, with particular reference to its origin from fibroadenomla

,X''e e ,Z' b*A;.t w;t

k7l: . ;NW et 4

I

i.0.

P.

*.

FIG. 11. Case 20, Table I. Adenosarcoma showing two very abnormal acini (centre) surrounded by sarcomatous tissuewhich closely resembles hyaline cartilage. (Haematoxylin and eosin x 140.)

'PURE' SARCOMAS

FIBROMYXOSARCOMA WITH DEATH FROM SARCOMATOSIS Awoman of 62 (Case 33, Table III) had noticed a lump inthe right breast 16 weeks previously. The lump hadslightly increased in size during the six weeks before shereported but she had no pain or discharge from thenipple. A lump was visible in the lower quadrant ofthe right breast below the nipple. To palpation it was thesize of a walnut, hard and craggy and fixed to skin butnot to muscle. No glands were palpable and the nipplewas not retracted. The left breast was normal. Radicalmastectomy was performed. The outer lower quadrant ofthe operation specimen contained a spherical mass 4 cm.diameter, which, though of hard consistence, showedirregular translucent gelatinous areas on cross section.The tumour adhered to the skin over an area about15 mm. diameter. No enlarged lymph nodes were found.Histologically, the tumour was a fibromyxosarcomainvading the surrounding breast and fatty tissues as wellas the subcutaneous tissue. There was extensive necrosis.

Noteworthy was the intense infiltration of the invadingedge of the tumour by eosinophil leucocytes. Post-operative radiotherapy was given. The patient died oneyear after the operation; necropsy showed sarcomatosis.

This is a more malignant form of breast sarcoma.Despite the comparatively small size of the primarygrowth, radical removal of the breast and radio-therapy failed to save the patient who died ofsarcomatosis within a year. All the same, it is note-worthy that no secondary deposits were found in theaxillary nodes in the operation specimen, despite theinvasiveness of the tumour.

OSTEOCHONDROSARCOMA WITH DEATH FROM SARCOMA-TOSIS Ten days before reporting, a woman of 52 (Case 37,Table III) discovered a lump in the left breast when wash-ing herself. She had no pain or discharge from the nipple.Her health had been good and her weight steady. Shehad two children and had had no lactation troubles. The

I1I

Vi

t.A sk..N, f: "'Ma..i ..4.0 .A.".

Ole

lw



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


TABLE III

'PURE' SARCOMAS


RelationtoMenopause


Histology

29 77 6 wk.

30 71 6 wk.

31 65 I yr.

32 62 6 wk.Breast inflamedand indurated

33 62 16 wk.

34 59

35 59 4 mth.

36 57 10 yr.Recent growthand ulceration

37 52 10 days

38 26 Present foryr. and grow-

ing for 4 mth.

39 3 mth.

After Upper half Ill-defined hae-morrhagic tumourabout 5 cm. diametercontaining severalsmall cysts

After Left Cystic tumour5 0x3 5x 1-5 cm.

After Right

After Inner leftupperquadrant

Large (20 cm. dia-meter) spherical, well-defined, softish mass,

with areas of hae-morrhage and cysticchangeSkin adherent andulcerated over tumour

5 cm. diameter

After Lower right Hard nodule (4 cm.

outer diameter) fixed toquadrant skin

Pleomorphic and very Radical mastectomyinvasive giant-cell sarcoma Died 8 months laterwith osteoid areas (no necropsy)



Pleomorphic sarcomawith giant-cell areas

Fibromyxosarcoma

After Left Over 2 5 cm. diameter Low-grade sarcoma

After Breast replaced by a

large, ulcerated, softovoid greyish mass

(19 x I I cm.)

After Left Breast greatly en-

larged and replaced bygreyish, gelatinousand necrotic tumour

After Lower leftouterquadrant

Ovoid firm mass,2-5 cm. diameterwith gritty parts on

section

Before Lower left Tough, pale brownhalf tumour about 4 cm.

diameter

Lower half Friable, only partlyencapsulated, 3-5 cm.

diameter

Fibromyxosarcoma withpleomorphic and giant-cell areas, particularly whereinvading fatty tissuesExtensive necrosis andblood vessel invasionPolymorphs and lympho-cytes numerous aroundperiphery

Myxosarcoma

Pleomorphic osteo-chondrosarcoma, withgiant cells and invasionof blood vessels

Anaplastic spindle-celltumour

Low-grade spindle-cellsarcoma

Local excisionLocal recurrence 9months later ofspindle-cell sarcomaRadical mastectomyNo recurrence 9 years

later

Local mastectomyAlive 21 months laterbut multiple secon-daries in bones andelsewhere

Local mastectomy andradiotherapyDied 6 years later withperhaps unrelatedsymptoms (nonecropsy)

Radical mastectomyand radiotherapyDied year later withdisseminatedmetastases (confirmedat necropsy)

Local excisionLocal recurrence 2years laterRadical mastectomyWell 3 years later

Local mastectomy

Not excisedDied 2 days afteradmission withoedema of lungsNo metastases atnecropsy

Radical mastectomyLung secondariesafter 14 months anddeath after 18 months

Local excision fol-lowed by radicalmastectomyDied I year 9 monthslater with dis-seminated metastases

Local mastectomyWell 2 years later

12

Treatment andPost-operative Course



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

Sarcoma of breast, with particular reJerence to its origin from fibroadenoma

we, VW

FIG. 12w ...

be.2=i t..__ q_t - .s 9. j .2.:

's r 597*_,v wT.s..... -s_Ff w ^:: *vZ _s.,.i.B<.

=_ # w__

FIG. 12. Case 36, TableIII. The very pleomor-phic tumour cells, somemultinucleate (centre),are forming abundantosteoid. In one area(below left of centre),this is calcifying. (Hae-matoxylin and eosin x400.)

FIG. 13. Case 36, TableIII. The tumour cellshave undergone cartil-aginous metaplasia.(Haematoxylin andeosin x 400.)

FIG. 13

13- --" 7,t.'Ji" --.I*1 W..*OF' lb.:

... 1.

.....

.77.1;

.9K.-

:.:



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


breast was of normal shape but contained an elongatedlump with lobulated surfaces in the lower outer quadrant.It was not tender or adherent to skin or to the deeptissues. No axillary or infraclavicular lymph glands werepalpable. Radical mastectomy was performed.The tumour was ovoid and measured about 2 5 cm. in

its long axis. It was partly cystic. The solid parts cut witha gritty sensation. There was no definite capsule. Histo-logically, the specimen was an osteochondrofibrosarcoma(Figs. 12 and 13). Cellular pleomorphism was great andmany giant cells, some resembling osteoclasts, werepresent. There were groups of tumour cells within manyblood vessels. Necrosis was extensive. Death fromsarcomatosis occurred 18 months after operation.

This is another example of highly malignant breastsarcoma: the tumour was small, it was noticed only10 days before operation, treatment was radical, butdisseminated secondaries were detected within abouta year of operation and death occurred a few monthslater. This outcome was readily forecast from thehistological features and particularly by the wide-spread vascular penetration by growth. The tumour'scomplex structure also illustrates the metaplasticpotentialities of breast mesenchyme.

DISCUSSION

AGE INCIDENCE

The ages of these 39 cases of sarcoma ranged from26 to 78 years, but the majority were in elderlywomen. Few long antedated the menopause, and theone which did do (Case 38, Table III) is muchthe least satisfactory of the whole series, in that thepatient, a woman of 26, had been exposed for someyears to oestrogens in the course of her employmentin a drug-making factory, and anaplasia of thetumour cells was so complete that the possibility ofher having spindle-cell carcinoma cannot beexcluded.

INCIDENCE

There is uncertainty over the frequency with whichsarcoma of breast occurs. Illingworth and Dick(1941) put it as high as 3% of all mammary tumours,whereas Willis (1948) diagnosed it only once among698 breast tumours. Our experience suggests that thetrue figure is somewhere between those extremes.Thus eight of the 39 cases occurred in the last 10years in one hospital, and during the same periodthere were 868 carcinomas of breast, that is, theeight sarcomas formed 0-9% of all malignant breasttumours. Undoubtedly, the widely varying estimatesof incidence are at least partly accounted for by thevagueness of the term 'malignancy'. Most breastsarcomas are of moderate malignancy by any

criteria, and since the large majority are curable byadequate surgery, a certain latitude in the inter-pretation of the histological features is permissiblewithout serious repercussions. It must be madeclear therefore that in all but a few of the presentseries of cases the features of malignancy were un-mistakable on microscopy; and the few which did notpossess such a histological structure (Cases 3 and 4,Table I) were malignant clinically. The larger matterof the histogenesis of sarcoma of breast also bears onthis question of incidence.

HISTOGENESIS

Probably the most important clue to the histogenesisof at least a significant proportion of breast sarcomasis provided by Table I which lists those tumourswhich undoubtedly developed from fibroadenomas,and by Table II, which lists the tumours with anadenosarcomatous structure. Fibroadenomas gener-ally arise in younger women and seem to behormone-dependent to a certain extent as mostof them tend to regress after the menopause.However, Table I shows that some fail to do so andstart to grow in an aggressive way. Although thisnew growth is primarily on the part of the fibro-adonoma's connective tissue, the epithelium is veryoften involved too, for the malignant parts of manyof the tumours listed in Table I contained glandularelements which appeared to be integral parts of thenew tissue, that is, they were adenosarcomas. Infact, in several cases the abnormal shape of theglands and the pleomorphism and mitotic activity ofthe epithelial cells (Fig. 5) suggested carcinoma. Butthe only certain example of carcinomatous trans-formation of the epithelial element was provided byCase 15 (Table I) where the axillary lymph-nodemetastasis contained both epithelial and mesen-chymal components (Fig. 8). The relationshipbetween breast epithelium and its supporting stromais very intimate, and the fibroadenoma is a true'mixed' tumour. It is hardly surprising therefore thatadenosarcomatous and even carcinosarcomatoustransformation should occur in a fibroadenoma ofbreast. What is remarkable is the way the epitheliumand connective tissue remain together, even to theextent of metastasizing to a lymph node (Fig. 8).As fibroadenomas occasionally undergo adeno-

sarcomatous change the adenosarcomas listed inTable II might well have originated in pericana-licular fibroadenomata. If they did not, one mustpostulate either that sarcomatous breast tissue has amarked tendency to induce the epithelium to growalong with it or that the unknown carcinogen pro-duces simultaneous malignant change in connectivetissue and epithelium.

14



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

Sarcoma of breast, with particular reference to its origin from fibroadenoma

With regard to the 'pure' sarcomas (Table LII),experience with the cases listed in Tables I and II

suggests that the more closely these are examined,the greater is the proportion found to containepithelial elements. Another feature which tends tolink this group with many of the tumours in TablesI and LI is their marked tendency to be myxosar-comatous.

Lastly, it is probably significant that breastsarcoma is a tumour of the 'carcinoma age' group.In this it is reminiscent of such 'secondary' sarcomasas osteogenic sarcoma developing in bones affectedby Paget's disease or of leiomyosarcoma arising froma uterine fibromyoma. By analogy breast sarcoma

may well be a 'secondary' type of growth.

TERMINOLOGY

The terminology of breast conditions is notoriouslyconfused and several other terms might be applied tosome of the tumours described in this paper. One is'giant' fibroadenoma. Fibroadenomas occur in allsizes, and, understandably, many pathologists havelabelled as 'giant' those tumours, pericanalicular and/or intracanalicular in type, which struck them as un-usually large. However, the term was given a more

specific meaning by Treves and Sunderland (1951)when they stated that the essential feature of a giantfibroadenoma is the disproportionate growth of themesenchymal component (Fig. 1); that is, the term'giant' should refer only to the relative size andcellularity of the intracanalicular 'clubs' or processes.

Naturally, most giant fibroadenomas are largetumours, but small ones (1 or 2 cm. diameter)possessing such giant processes do occur. There ismuch to be said for Treves and Sunderland's usage,

but unfortunately there is little doubt that patholo-gists will continue to use giant to describe the size ofa tumour and even to apply it to unusually largepericanalicular fibroadenomas.

Giant fibroadenomas in the strict sense of theterm may be simple or malignant and the malignantones form a substantial fraction of all breastsarcomas. Ten of the 16 tumours in Table I qualifyas giant fibroadenomas, and these therefore con-

stitute about 25% of the present series of 39 sar-

comas. Six of these 10 cases occurred in 20 years in

one of the three hospitals concerned in the presentstudy. During the same period there were in the samehospital 11 other patients with tumours which were

histologically benign or of doubtful malignancy andwhich are therefore not included in Table I. Inci-dentally, the reasons for not including at least some

of these 11 neoplasms in Table I must be arbitraryand personal to a certain extent; but at least thepatients have had no subsequent trouble after only

local excision of the tumour. These figures also givesome idea of the frequency with which giant fibro-adenomas are encountered and of the relative pro-portions of benign and malignant forms; but thefigures are very approximate indeed, since it isobviously impossible to lay down criteria, clinicaland/or histological, which will readily distinguishbetween the benign and malignant forms of giantfibroadenomas, and it is wise to regard all withsuspicion. Fortunately, with adequate surgery, mostwill not justify these suspicions, and even anenormous ulcerated specimen may fail to providehistological or post-operative evidence to justify adiagnosis of sarcoma (Goodall and Curran, 1953).

Cystosarcoma phyllodes is an old term (Muller,1838) which still confuses surgeons and pathologists.Muller originally applied it to tumours which showeda leaf-like or cabbage-like pattern when cut across.This characteristic appearance results from greatovergrowth of the intracanalicular processes of afibroadenoma, so that large fibromyxomatousmasses form which are separated by epithelium-lined, slit-like crevices. Thus the term is roughlysynonymous with the giant fibroadenoma of Trevesand Sunderland, except that the intracanaliculargrowths should be sufficiently large and numerousfor the characteristic leaf-like pattern to be readilyvisible to the naked eye. Muller did not assess themalignancy of his tumours histologically but reliedonly on their gross appearance and clinical behaviour,and so there was no correlation with microscopicstructure. It is not surprising therefore that his termhas been used very loosely and applied to tumoursof widely varying histological appearance (forreview, see Goodall and Curran). Perhaps the mostuseful feature of the term lies in its ominous sound,which serves as a warning that tumours with thisnaked-eye structure should not be treated toolightly! However, only a minority of breast sarcomasare 'cystosarcomatous' and the majority of 'phyl-lodes' type tumours behave benignly. In otherwords, the term is a source of confusion and itshould be discarded. So, incidentally, should theterm 'Brodie's serocystic disease', which is equallyvague.

TREATMENT AND PROGNOSIS

It is noteworthy that of the 17 patients who hadradical mastectomy either at the initial operation orfor a recurrent growth, only one had secondarydeposits in the axillary lymph nodes, and this singleexception was a very unusual tumour indeed, acarcinosarcoma (Case 15, Table I). This fact prob-ably illustrates the reluctance of sarcomas to use thelymphatic route for metastasis and suggests that

15



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


simple mastectomy is sufficient treatment for sar-coma of breast. There seems to be little point, there-fore, in doing a radical mastectomy unless theepithelial elements in the tumour appear to befrankly carcinomatous, an occurrence so uncommon(Case 15, Table 1) as to be discounted when con-sidering treatment. Simple mastectomy does not, ofcourse, eliminate the possibility of later blood-borne metastasis, since even radical mastectomyfailed to save several women from death from dis-seminated metastases. By and large, however, breastsarcoma does not metastasize readily and even whenit does the metastases may be solitary and resectable(Case20, TablelI). On the other hand,metastasismayappear after a considerable interval (Case 4,Table I).Although radical mastectomy is therefore probablynot necessary, inadequate primary operations mustbe avoided, for a considerable proportion of thosecases treated by local excision had local recurrences(for example, Cases 19 and 22, Table II), sometimesafter several years (Case 10, Table I).

PARITY AND MARITAL STATE

On the incomplete data available, neither of thesefactors seems to have a connexion with the appear-ance of breast sarcoma.

ASSESSMENT OF MALIGNANCY

As already stated, the fact that most breast sarcomasare curable by adequate surgery must account inpart at least for the wide variation in the frequencywith which the diagnosis of sarcoma is made, sinceretrospective correction of the diagnosis is rarelyrequired. Even with care, however, histology is notan infallible guide with the types of tumour listed inTables I and II. Thus several tumours provedclinically malignant despite the pathologists' in-ability to demonstrate, even in retrospect, a firmhistological basis for their behaviour, the moststriking example being provided by Case 4 (Table I),a patient who died from sarcomatosis four yearsafter mastectomy for a large 'fibroadenoma' (Collinsand Dodge, 1959). Several reasons for this dis-crepancy can be suggested. Where the tumour hasdeveloped from a fibroadenoma, only a part of thefibroadenoma's 'stroma' may have undergonemalignant change, and this part may be readilymissed when relatively few blocks are taken forhistology. Many of these sarcomas invade the tissues

on a very broad front and often acquire a dense butincomplete fibrous 'capsule' (Figs. 4, 9, and 10)which can mislead the careless observer. In additionbreast sarcomas tend to be noticeably mucoidmacroscopically and myxomatous histologically; andmyxomatous tumours are notoriously unpredictable.Willis (1948) believes, in fact, that it is not worth-while trying to distinguish between myxoma andmyxosarcoma and that all prominently myxo-matous tumours should be regarded as malignant.Probably the most important factor of all, however,is the tendency for the pathologist to regard asinvariably benign anything even vaguely resemblinga fibroadenoma and to ignore quite definite histo-logical evidence of malignancy. 'Pure' sarcomas(Table III) give him little trouble, on the otherhand, although he may fail to realize that asmall minority of breast sarcomas are very invasiveand are able to kill within a remarkably short periodof time.With regard to prognosis, the significance of bone

or osteoid in the tumour is uncertain. Admittedlythey were present in two of the three rapidly-fatalcases (Cases 29, 33, and 37, Table III) but they werealso present in four other cases (Case 9, Table Iand Cases 18, 24, and 25, Table II), of which onlyone has had a reasonably long follow-up (Case 9,Table I), so far with no signs of recurrence. Asmight be expected, invasion of blood vessels wasprominent in those tumours which exhibited thegreatest degree of clinical malignancy (e.g., Cases 29and 37, Table III).There is no apparent connexion between prognosis

and either the size of tumour at operation or thelength of time it had been known to be present beforeoperation.

We are indebted to the clinicians of the Royal Infirmariesof Glasgow and Sheffield and of St. Thomas's Hospital,London, for allowing us to use their records; to Dr.Van der Merwe of the General Hospital, Burton-on-Trent, for Case 5; and to Miss J. Read for secretarialhelp.

REFERENCES

Collins, P. G., and Dodge, 0. G. (1959). Brit. J. Surg., 47, 37.Goodall, A. L., and Curran, R. C. (1953). Ibid., 40, 479.Illingworth, C. F. W., and Dick, B. M. (1941). A Text-book ofSurgical

Pathology, 4th ed., p. 382. Churchill, London.Muller, J. (1838). Ueber den feinern Bau und die Formen der krank-

haften, Geschwulste., p. 56. Reimer, Berlin.Treves, N., and Sunderland, D. A. (1951). Cancer, 4, 1286.Willis, R. A. (1948). Pathology of Tumours, 1st ed., pp. 210 and 654.

Butterworth, London.

16



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

Technical methods

ItsA method for obtaining concentrates

of eosinophils from bloodR. F. ALEXANDER AND A. 1. SPRIGGS Fronm the

Churchill Hospital, Oxford

'p

0

FIG. 3. Examples of finished preparations, stained with0-5% azocarmine.

SUMMARY

A micro agar-gel precipitation technique is described inwhich the reagents are applied to the agar by means of ablock of perspex through which holes have been drilled.The method allows a high degree of precision in placingthe reagents on the agar with consequent good repeata-bility of results.

CORRECTIONS

We very much regret that Figs. 2, 4, and 7 of the paper'Sarcoma of breast, with particular reference to itsorigin from fibroadenoma' by R. C. Curran and 0. G.Dodge (J. clin. Path., 15, 1-16) have been printed upsidedown. In the legends accordingly please read 'right' for'left' and vice versa. Also Figs. 12 and 13 refer to Case 37,not to Case 36 as printed.

Professor N. H. Martin asks that the last phrase of thelast sentence of the first paragraph of the Discussion inhis paper 'Serum sialic acid levels in health and disease'(J. clin. Path., 15, 71) 'whereas an increase in the a fractionwould have little effect on this ratio', be deleted.

In a previous publication (Spriggs and Alexander, 1960)we described an albumin gradient method for separatingthe different white cells of blood. This was applied to theisolation of tumour cells but it was also noted that verypure suspensions of neutrophil polymorphonuclearscould be obtained by the same method. It has now beenfound that the eosinophil leucocytes can be collected bythe albumin method in a separate layer, and if the bloodsample comes from a patient with eosinophilia it is some-times possible to pipette these cells off in a high degreeof purity.The albumin gradient method need not be described

again, but it should be noted that all equipment must besiliconed. The cells form a series of layers according totheir specific gravity, as shown in Fig. 1. The platelets

-.f,platelets and monocytes

.* .* lymphocytes (inconstant)

,r';.;,t .e. neu troph i I s~.-

red cells....: ; cosinophils

FIG. 1.

are on top, with most of the leucocytes below them, andany residual red cells not removed by the preliminarysedimentation are lower still. When eosinophils are

Received for publication 20 July 1961.

188

0O ... I*

Y. 11'k:: ..:0W.7

sarcoma of breast, with particular reference to its origin ... · fibroadenoma with cellular...

Documents