The Surgeon and the Pathologist

Address.THE SURGEON AND THE PATHOLOGIST. A

PLEA FOR RECIPROCITY AS ILLUSTRATEDBY THE CONSIDERATION OF THE CLASSI-FICATION AND TREATMENT OF BENIGNTUMORS OF THE BREAST.*

BY J. COLLINS WARREN, M.D., BOSTON.

SYNOPSIS.1. Introduction: Subject of paper. Confusion of

nomenclature. {National and local differences.Differences between clinician and pathologist.Need of co-operation. Clearing house.

2. Material: 658 consecutive cases of chronic diseaseof breast in M. G. H.; 100 cases from privatepractice (70 diagnoses in 199 cases of benign breasttumor in M. G. H.).

3. Classification: Normal anatomy of breast. Rela-tion to tumors and diseases. Classification offered.Variation from other classifications.

4. Carcinoma: 517 cases. Proportion to other tumors,68 % (70% M. G. H.) (58 % J. C. W.). Varieties.Age incidence. Occurrence with or resulting frombenign tumors and diseases, as Involution, papil-lary cyst adenoma, periductal fibroma, Paget'sdisease.

5. Fibho-Epithelial Tumors: Fibrous Group. Peri-ductal tumors. Fibroma, myxoma, sarcoma.Frequency. Pathology. Clinical symptoms.Diagnosis. Course. Prognosis. Treatment.

6. Fibro-Epithelial Tumors: Epithelial group. Cystadenoma. Fibro-Cyst-Adenoma : Rare. Epi-thelial exaggeration of periductal fibroma. Patho-logic symptoms. Treatment. Papillary Cyst-Adenoma: Pathology. Symptoms. Prognosis.Treatment.

7. Hyperplasia. Diffuse Hypertrophy: (2) Vari-eties. Rare. Relation to pregnancy and to geni-tal abnormalities. Pathology. Symptoms.Treatment.

8. Hyperplasia. Abnormal Involution. " ChronicCystic Disease." Historical. Synonyms. Path-ology. Cystic and proliferative. Types of pro-liferation. Symptoms. Course untreated. Prog-nosis. Treatment.

9. Chronic Inflammation: Chronic abscess. Ductalmastitis. Relation to lactation. Symptoms.Course. Pathology and Treatment. Galactocele.

10. Single Retention Cyst: Mechancial origin. Rarity.Symptoms. Confusion with abnormal involution.

11. Non-Indigenous Tumors of Breast: Sarcoma." Solid sarcoma." Rare. No relation to breast.Symptoms. Diagnosis. Treatment. Lipoma.Retro-mammary. Symptoms and treatment.Lymphangioma.

12. Supernumerary Breasts: Occurrence. Situation.Tendency to tumor formation.

13. Treatment of benign tumors qf the breast. "Plasticresection." Indications. Method. Results.

In selecting a subject in a special departmentof medicine which should be of interest to thegreat body of members of this Association, it hasoccurred to me that the discussion of the diseasesof the mammary gland would be appropriate,as the classification of the diseases of this organis at present in a more or less unsettled state,and the operative treatment of many of the be-nign lesions of the breast is a subject which isstill open to discussion.

The great impetus which laboratory work hasreceived in recent years in the various medical

centers of this country is evidence that the interesttaken in the scientific side of medicine has pro-gressed steadily since the awakening of modernpathology in the middle of the last century.American physicians have proved themselvesapt pupils, and the beginnings of the generationwhich is now passing off the stage have alreadyborne fruit which one cannot regard otherwisethan with satisfaction. The time has alreadyarrived when we in this country are no longercontent to receive the dicta of science at secondhand, but feel quite competent to take the ini-tiative on ourselves, as our many substantialcontributions to medical science amply show.Indeed, I feel convinced that the very practicalside of our national character is a most valuablequality in enabling us to unravel some of themore complicated questions which have beenleft unsolved by the foreign scientists. Thelaboratory workers' range of vision, however,is often at best but microscopic, and it is nowonder that, unaided, he sometimes misinterpretsthe changes rung by nature on some simpletheme. Without the laboratory worker's aid,however, the clinician becomes swamped in amass of clinical detail from which alone he hasagain and again been unable to extricate himself.There can be little question that the combinationof energies which accomplishes most in surgicalprogress is that obtained by the co-operation ofthe laboratory investigator with the surgeon ofclinical experience. Too long have these twodepartments of medicine conducted themselvesindependently and, as I feel, greatly to the dis-advantage of them both. The clinician and thelaboratory man, however, in these days of spe-cialism seem sometimes in danger of driftingeven farther apart, and we find some of the mod-ern problems sadly in need of a pathologic " clear-ing-house " to put things on a more practicalbasis. This, it seems to me, is the moral whichmust be read between the lines of the paperwhich I have the honor to present to you thisevening.

In no department of surgery has the classifi-cation of the diseases of an organ or the patho-logic nomenclature been more confusing than inthe case of the diseases of the mammary gland.Not only has the nomenclature of the diseasesof the breast been subject to that continual re-vision to which all medical terms are liable withadvancing knowledge, but in this case nationaland even local systems have added again to theconfusion. The French school, the German,the English, and the American schools of medi-cine rarely agree in their description of anytumor of the breast other than carcinoma, andin the case of the inflammatory or involutionchanges of the breast the variations in nomen-clature are notorious. In view of the alwayshorrible possibility of cancer, however, no womancan regard a lump of any size, shape, or consis-tency in the breast with equanimity, and in no

organ does the presence of disease in its earlierstages furnish fewer data for accurate diagnosis.The cry for aid and comfort is always an urgent

*Oration on Surgery at the Fifty-sixth Annual Session of theAmerican Medical Association, Portland, Ore., July 11-14, 1905.Published through the courtesy of the Journal of the AmericanMedical Association.

The Boston Medical and Surgical Journal as published by The New England Journal of Medicine. Downloaded from nejm.org at NEW YORK UNIVERSITY on October 11, 2014.

For personal use only. No other uses without permission. From the NEJM Archive. Copyright © 2010 Massachusetts Medical Society.

one, and it behooves the surgeon, therefore, andthe general practitioner as well, to be in a posi-tion to meet the demand for accurate informationwith an intelligent and harmonious reply. Sucha statement can be furnished only by bringingthe modern views of the pathology of thesediseases into focus with the clinical pictures whichthey present. In the hope that I may pavethe way for the clinician and the pathologist to"get together " and to put the classification ofdiseases of the breast on a more intelligible andpracticable foundation, I have selected this fieldof surgery for your consideration.

The material which has served as a basis ofthis investigation includes, in all, 758 cases ofdiseases of the breast; 658 cases were takenfrom the records of the Massachusetts GeneralHospital.1 The diseases of the male breast w-ere

purposely omitted from this table of cases, andof the total number of cases of diseases of thefemale breast 10 were omitted because of insuffi-cient data to establish a diagnosis. Acute mam-

mary abscess was also excluded from considera-tion. With these exceptions all cases of diseaseof the mammary gland are included which enteredthe Massachusetts General Hospital during thedecade from 1895 to 1905. One hundred addi-tional consecutive cases of disease of the breasthave been obtained from my own private prac-tice, making up the total of 758 cases of all varie-ties with which we have to deal. Pathologicreports were available in nearly every case, andin many instances the actual specimens couldbe obtained and were used for further study andfor purposes of illustration. Four hundredand fifty-nine of the hospital cases and 58 ofthe cases in private practice were cancer of thebreast. These were included for purpose ofcomparison and in order better to establish thesignificant data for diagnosis. The greater partof the study, however, has been devoted to thebenign diseases of the breast, 42 in private prac-tice, 199 in the hospital series

—

a total of 241cases of benign diseases of the breast. Theseare exclusive of re-entries or of cases in whichtwo diseases of different types occurred. As an

instance of the confusion in nomenclature, Imight here state that in 199 cases of benigndisease of the breast occurring in the Massachu-setts General Hospital in ten years, 70 differentpathologic diagnosis appeared on the recordbooks.

Before taking up the question of classificationof the benign tumors of the breast, I would liketo call your attention to some anatomic peculiar-ities of the mammary gland which distinguishit from any other organ in the body. At notime during the various periods of its life historydoes it have an enduring type of anatomic struc-ture. At birth the gland is represented by a

series of radiating ducts with club-shaped ex-

tremities, but with no well-developed acini.The epithelium of these ducts is often in a stateof active proliferation, causing swelling and ten-derness of the gland, with ectasia of the ductsand supposed inflammation (acute mastitis of

infants). Infection through the nipple mayundoubtedly occur at this time, but in the vastmajority of cases the swelling and tendernessare due entirely to the developmental activityof the gland. No material change in the glandtakes place from this time until the age of puberty,when the development of the acini first reallybegins. An enlargement of the breast, withdevelopment of considerable tenderness is oftennoticed at puberty in males as well as females.The epithelial structures are now beginning tothemselves prominent, but there is also a char-acteristic change in the tissue immediately aroundthe ducts. The periductal connective tissuebecomes transparent and rich in nuclei and pre-sents a strong contrast to the interstitial tissueof the gland. This transparent periductal tissueis peculiar to the mammary gland. It givesthe familiar consistency and shape to the virginbreast and enters largely into the new growthsof the breast which take origin at this stage ofits life history.

At the time of pregnancy another great changetakes place in the anatomy of the breast. Nowthe epithelial activity of the gland is at its height,the acini multiply enormously, the tree may besaid to be in full leaf. Meanwhile the interstitialtissue becomes stretched and the periductaltissue becomes less prominent. After lactationmany of the acini wither and disappear, theinterstitial tissue is relaxed and the breast be-comes pendulous.

The declining period in the life history of themammary gland may begin in middle life, andpresents a condition which is of great interestto the surgeon. The breast, as it were, graduallydries up. Change is noticed especially in theacini and in the interstitial tissue. The epithe-lium is no longer in a state of active proliferation,but degenerative changes gradually gain the upperhand. Some ducts are choked with epithelialdébris, while others are compressed by the con-

tracting interstitial tissue of the gland, and theblood vessels become thrombosed and disappear.Obstruction takes place here and there, andslight dilatation of the small ducts and acinioccurs, with induration of the surrounding tissues.If the breast in this condition is gently pressedagainst the thorax with the flat of the fingersit will be found to contain a large number of smallhard nodules. This condition I have describedas the " cobble-stone " feel of the breast. Itis present in almost every breast at a certainstage of its decline, and should be regarded as anormal condition at this period of the life historyof the gland. If the gland is incised these noduleswill be found to be composed of fibrous tissueor of small cysts filled with a slightly discoloredfluid which derives its hue from solution of theold blood crystals and pigment from the throm-bosed blood vessels.

As the gland tissue is absorbed after lactation,2 3

and with advancing years, a certain amount ofadipose tissue takes its place, and in old age wefind the mammary gland represented merely bya few ducts near the nipple and some small strands



of fibrous stroma, the meshes of which are wellfilled with fat. Keeping in mind the anatomicpeculiarities of the mammary gland, we findtwo distinct structures intimately associatedwith different periods of its life history. I referto the periductal fibrous tissue so prominentin early life and to the epithelial elements whichat certain later periods, as in lactation, attainthe perfection of their development. Thesetwo structures are so intimately associated onewith another that it is impossible for a neoplasticformation to develop without partaking some-what of the characteristics of both tissues. Theperiductal fibrous tissue of the gland cannotform a tumor without involving, to a certainextent, the epithelium of the ducts. Nor canthe epithelial elements grow entirely independentof their fibrous support without placing thegrowth at once in the category of carcinoma. Aconsistent classification of the tumors of thebreast must, therefore, take into considerationboth of these elements, and the term " fibro-epithelial " tumors, suggested by Ribbert,4 bestmeets these indications.

Tumors are classified, however, either accord-ing to the nature of the tissue of their origin orthe tissue which forms their most significantpart. In these fibro-epithelial tumors, therefore,certain subdivisions are possible and, indeed,important, according to the tissues which aremost seriously involved. A ready distinctionis suggested at once between the tumors whichare chiefly of a fibrous character and those inwhich the epithelium is more involved. Thisdistinction, moreover, has an important clinicalvalue, as there is a tendency of the fibrous grouptoward sarcomatous change and of the epithelialgroup toward carcinomatous degeneration. Underfibro-epithelial tumors of a fibrous character we

may thus distinguish three divisions—

the peri-

ductal fibroma, periductal myoxma, and peri-ductal sarcoma. Those of the epithelial groupwe can divide into the fibro-cyst-adenoma andthe papillary cyst-adenoma. In the tumors ofthe fibrous group the fibrous tissue is the pre-dominant feature of the tumor, epithelium beingpresent, but playing only a subordinate part.In the epithelial group the fibrous tissue is presentalso, but merely in sufficient quantity as to beregarded as a stroma, the growth of epitheliumbeing the predominant feature of the tumor.

The term hyperplasia I have reserved for thosediffuse changes which are characteristic of thebreast at certain periods of its life history. Fromtheir very diffuse character, these cannot properlybe placed within the category of tumor formation.They cover not only that common conditionwhich we find so characteristic a feature of theinvolution stage of the normal breast and whichhas passed under the many terms of " chroniccystic mastitis," " cystic disease of the breast,"etc., but also that rare diffuse hyperplastic growthof all the elements of the gland at the period ofhighest physical development

—

" diffuse hyper-trophy." The chronic inflammatory processesof the mammary gland present no special diffi-culties in classification; but they, as well asthe non-indigenous tumors of the breast, areincluded in the table to round out the classifi-cation of all lesions, with the exception of acutemammary abscess.

The diagnosis of diseases of the breast is un-

doubtedly the first consideration, and it is forthis reason that the cases of carcinoma havebeen included in the list below and must nowreceive a brief analysis.

Cancer occurred in 459, or 70%, of all tumorsin the hospital cases and in 58, or 58%, of thecases hi private practice. The proportion ofcancer to other breast tumor was thus greater

Table I—

Classification of 758 Cases of Breast Tumor.Private Per

M. G. H. Cases. Total. Cent.

Carcinoma. 459 58 517 68Fibro-EpithelialTumors.

...

(86) (11)(1) Fibroustype. ... (70) (9)

1. Periductal fibroma. 48 6 54 72. Periductal myxoma.'.

.

10 2 12 1.63. Periductal sarcoma. 4 0 4 .5

(2) Epithelial type (Cyst adenoma). ... (16) (2)1. Fibro-cyst-adenoma. 3 1 4 .52. Papillary cyst-adenoma. 9 3 12 1.6

Hyperplasia.1. Diffusehypertrophv. 2 0 2 .22. Abnormal involution. 87 28 115 15

(Cystic, (59)(Proliferative, 56)

Chronic Inflammation. ... (28) (3.7)1. Eczema ofnipple. 2 0 2 .22. Chronicabscess. 5 0 5 .63. Due talmastitis. 3 0 3 .34. Tuberculosis. 13 0 13 1.75. Single rentention cyst. 5 0 5 .6

Non-indigenoustumors. ... (9) (1-2)1. Sarcoma. 4 0 4 .52. Lipoma. 3 1 4 .53. Lymphangioina. 1 0 1 .1

Supernumerarybreast. 1 1 .1

658 100 758



in hospital cases than in private practice. Thisis undoubtedly due to the fact that the mildercases of benign tumor are occasionally treatedin the out-patient department of a hospital and,therefore, do not enter into this number of cases,whereas in private practice all cases are included.The fact, however, is obvious that cancer formsthe great majority of actual new growths of thebreast, and if the conditions known as abnormalinvolution are left out of consideration thispredominance of cancer over other tumors be-comes even more conspicuous.

Paget's disease of the nipple occurred in thisnumber of cases six times in connection withcancer of one or another variety; and eczemaof the nipple and areola, similar in every way toPaget's disease, but without evidence of malig-nancy, was found in two additional cases duringthis ten-year period.5

The forms of cancer which have developedfrom other benign tumors or conditions of thebreast are of special interest to us in this con-nection. During this period thirteen cases oc-curred in the hospital and four in the cases of

Table II—

Incidence of

No. of 10-20Cases. Yrs.

Carcinoma. 459 ....

Periductal sarcoma. 4 ....

Periductal myxoma. 10 ....

Periductal fibroma. 48 '445%

Fibro-cyst-adenoma. 3 ....

Papillary cyst-adenoma. 9 ....

Diffuse hypertrophy. 2 111%

Abnormal involution. 87 ...

Eczema of nipple. 2....

Chronic abscess and ductal mastitis 8 ....

Single retention cyst. 5 ....

Tuberculosis. 13 333%

Non-indigenous sarcoma. 4 ....

•

Other non-indigenous tumors. 4 111%

Total Cases. 658 9

Breast Tumors by Age Decades.20-30 30-40 40-50 50-60 60-70 70-80Yrs.

612%

510%2650%

1

713%

24%24%36%

12%

Yrs.6760%

11%

1311%

22%11%

2018%

22%11%33%22%

Yrs. Yrs.151 13070% 85%

1 3 4% 2%

31.5%42%

31.5%1•4%

4521%ï%31.5%21%42%1•4%

1•6%•6%

21%

138%2

1•6%

Yrs.8092%

33%

22%

6%

11%11%

80Yrs. Yrs.24 1

100% 100%

53 112 218 154 87 24 1

private practice, in which cancer resulted from,or was present in connection with, benign dis-eases of the breast.

The involution changes of the breast are thosewhich accompany or precede cancer more fre-quently than any other form of benign lesion,and by far the greatest number of cases of thischaracter show the picture of adeno-carcinoma.Nine cases of adeno-carcinoma and three ofmedullary carcinoma occurred with abnormalinvolution. Three cases of scirrhus cancer alsooccurred which presented the lesions of abnormalinvolution. In the hospital series also two cases

of adeno-carcinoma occurred in connection withtumors of the papillary cyst-adenoma type.The significance of these cases of cancer will beespecially considered under the discussion of thebenign tumors of the breast.

The age incidence of the different tumors anddiseases of the breast is a matter of considerableimportance in diagnosis. I have arranged Table2 to show the number and percentage of casesof the different lesions occurring in each decadeof life.

By this table it is to be seen that the proportionof cancer to other tumors of the breast is a con-

stantly increasing one to the end of life. Wrhilebetween twenty and thirty years of age canceroccurs in only 6% of all cases, after seventy yearsthe percentage is 100. All of the twenty-fivetumors of the breast occurring in the Massa-chusetts General Hospital in ten years, inwomen who were over seventy years of age,were cancer.

The periductal fibromata, on the other hand,are most numerous in the third decade fromtwenty to thirty, and constitute 50% of all

tumors occurring at this period. From this timethey gradually diminish in frequency until thesixth decade, after which they disappear.

Of the cyst-adenomata, we find that tumor(fibro-cyst-adenoma) which is most nearly relatedto the periductal fibromata, appearing in thesame decades, the third and fourth, with themajority of the latter, while the. papillary cyst-adenoma, beginning in the fourth decade, con-tinues to the seventh, and, therefore, belongs to a

distinctly later period of life.Abnormal involution reaches high-water mark

in the fifth decade, forty to fifty years, with 21%



of the tumors occurring in this period. Thiscorresponds closely with the average age givenfor abnormal involution, which is from forty-twoto forty-five. The age incidence of other dis-eases is of less significance. The inflammatoryprocesses, as might be expected, occur especiallyin the early decades, and tuberculosis, in thisseries of cases, appears to have affected the breastsof younger women than has generally been be-lieved. The non-indigenous sarcomata of thebreast, though few in number, occur rather in themiddle period of life.

FIBRO-EPITHKLIAL TUMORS.

The fibro-epithelial tumors now present them-selves for consideration. These new growths ofthe breast have been described under so manydifferent names that even an enumeration of alltheir titles is impossible. Before the middleof the last century all tumors of the breast wereregarded as cancer. When Cruveilhier firstrecognized the existence of benign tumors ofthe breast a storm of opposition was raised in theFrench Academy of Medicine. Cruveilhier de-scribed a fibroma of the breast, but this namewas not accepted, and a multitude of othernames were used to describe the tumors of thischaracter. Thus Johannes Müller adopted theterm of cysto-sarcoma-phyllodes; Brodie, sero-cystic sarcoma; Sir Astley Cooper, hydatidtumor; Billroth,6 fibroma and cysto-sarcoma;Paget, chronic mammary and proliferous cysts.Adenocele, adeno-fibroma, fibro-adenoma, cyst-adenoma proliferum and intracanalicular-papil-lary-fibroma are all terms which have been usedat one time or another to apply to fibrous tumorsof the fibro-epithelial group. The reason forthis diversity of nomenclature is readily foundwhen the attempt is made to reduce the tumorto its component parts. In one case fibroustissue predominates, but in another the glandularstructures play a more important part. Ribbert4undoubtedly reaches the most satisfactory solu-tion of this difficulty by grouping the tumorsof this nature apart, not as fibroma or as adeno-ma, but as fibro-epithelial tumors partakingof the character of both tissues. One fact, how-ever, is apparent in the consideration of thesecases; and, although long ago pointed out byBillroth,6 this fact has not received the attentionof surgeons and pathologists which is its due.I refer to the fact that the chief constituent ofthe tumors of the fibrous type is the peculiartransparent periductal tissue of the female breast.This periductal tissue makes its appearance anddevelops during the period after puberty andbefore lactation, and it is at this period that thevast majority of tumors of the fibrous type occur.No one can deny that this periductal tissue isessentially a part of the glandular structure ofthe breast, and its close relation to the epitheliumof the ducts and the interdependence of the twotissues make it practically impossible for the oneto undergo the changes of tumor formation with-out the other participating in the process to a

greater or less extent. For these reasons I have

recently adopted the term periductal fibroma asthe most accurate name to apply to the tumorsof this character, and this term I venture to offerin the hope that the name " adeno " may bereserved for the tumors in which the epithelialelements play a more important part. Cysticdilatation of the ducts or of the characteristicclefts of the periductal fibroma may occur incertain instances, and is due in all probabilityto the obstruction of the pre-existing ducts.These cysts are thus secondary in character andof relatively secondary importance. The clinicalcharacteristics of the more cystic forms of peri-ductal fibroma are as little to be distinguishedas their pathological ones, and although the at-tempt was made in this investigation to separatethe cystic from the more solid tumors of this class,it was soon apparent that no data of value couldbe obtained. These tumors of the fibrous typemay, however, be divided according to theirrichness in cells or the character of their fibroustissue, into three groups, fibroma, myxoma, andsarcoma.

PERIDUCTAL FIBROMA.

Under this heading are included the true en-

capsulated fibro-epithelial tumors arising fromthe periductal fibrous tissue and composed chieflyof this tissue with a certain admixture of epithe-lial elements. Of this type of tumor there were48 cases in the hospital series and 6 in privatepractice, 7% of the whole number of diseasesof the breast. They varied greatly in size, froma bean to a cocoanut, and 15 cases had multipletumors in one or both breasts. The average sizewas about that of an English walnut. Thesetumors presented themselves clinically as hardnodules, freely movable in the breast tissue andunder the skin. In gross appearance they pre-sented firm, encapsulated tumors of a whitish,glistening appearance on section, and traversedby many cleft-like cavities, which were rarelydistended to such an extent as to justify the termof cysts. Microscopically they showed a firm,fibrous capsule limiting them from the rest of thebreast substance; but, while some were composedof dense tissue, others were looser in arrangement,and some had even a myxomatous appearance,which was attributable to edema.

The clinical features of tumors of this groupcould be summarized as follows :

—The greater number occurred in women fromtwenty to thirty years of age. Unmarriedwomen were most commonly affected. Thetumors were of slow growth and painless, althoughin certain instances increased sensitiveness atthe time of catamenia was recorded. The situa-tion of the tumors in the breast was variable,but the upper outer quadrant was most com-

monly involved. Amputation was required threetimes on account of the size of the tumor; in theother cases excision or resection was the operationwhich was performed. Eight of the 48 cases

showed cyst formation to a greater or less extent,but no clinical features served to distinguishthem from the other cases. In one case the



changes of abnormal involution were present inaddition to the fibroma, and in one instance alsocarcinoma appeared to have originated in a tumorof this character.

PERIDUCTAL MYXOMA.

Ten cases occurred in the hospital series andtwo in private practice. The myxomata wereobviously derived from the tumors of the peri-ductal-fibrous group and differed from them onlyby their size, which was somewhat greater, andby the gross and microscopic appearance ofmyxomatous tissue on section of the tumor.It seems probable, as recorded by a number ofother writers,7 8 4 that the myxomatous char-acter was due for the most part to the edema ofthe fibrous tissue which is characteristic ofthe periductal fibromata of the breast. Thesetumors occurred later in life than the fibromataand were of long duration. Rapid growth, how-ever, was not unusual and the tumors were gen-erally of large size. They were all hard in consist-ency. Necrosis and ulcération of the skin oc-curred in two cases, and enlarged axillary glandswere sometimes found. Adhesion to the skinwas present in one case in addition to two whichshowed ulcération. The pathologic character-istics were those of a large lobulated, well encap-sulated tumor showing cyst cavities or clefts,and a homogeneous transparent stroma. Exceptfor their size, it is obvious that these tumorsdiffered in no marked essentials from the tumorsof the fibroma class.

PERIDUCTAL SARCOMA.

Four cases of periductal sarcoma occurred inthis entire series. These were tumors of the fi-brous group with a stroma so cellular as to demandclassification as sarcoma. All, however, pre-sented the combination of cellular fibrous-tissuestroma and epithelial gland ducts. They were,as a rule, large, hard tumors involving the wholebreast, the size of a cocoanut or a child's head.They were lobulated and encapsulated, with amarked tendency to ulcération and to cyst for-mation. The skin above the tumors was, as arule, reddened and contained dilated veins, butwas not invariably adherent. Axillary glands,as a rule, were not involved. The tumors wereof long duration and the patients of advancedage, forty-four, fifty-one, fifty-three, and fifty-seven years. They were all married women.The growth of the tumor was slow at first, butin two cases had been more rapid within the lastfew months. Pain was insignificant. Ampu-tation was done in every case.

Periductal sarcoma is thus seen to be merelyan exaggeration of the periductal fibroma andmyxoma, occurring somewhat later in life andpresenting a tumor of greater size. Tumorsof this nature, although generally described assarcoma, have long been recognized to be of avery slight degree of malignancy. Schimmel-busch 7 8 has perhaps taken the most advancedposition in this regard. The encapsulation andthe absence of axillary or pulmonary metastasis

would appear to support this contention, and itmust be admitted that fibroma, myxoma andsarcoma are all tumors which are very closelyrelated. It is unusual, in fact, to find a specimenof periductal sarcoma which does not show insome portions the characteristic picture of myx-oma and of fibroma. The tendency to malig-nancy of the tumors of this group is only slight,and they cause disturbance principally of a localcharacter. The danger of carcinoma is very re-mote. They cannot, however, be regarded asinnocuous because of the limited tendency whichthey present to excessive growth and ulcération.For these reasons, operative removal is indicated.This may best be done in the small fibromataby the operation later to be described, underthe term of " plastic resection of the mammarygland." The larger sarcomata and myxomata,however, require amputation. Although thedetermination of the end results of periductalsarcoma has not been worked out, a point notcoming within the scope of this paper, my own

experience has led me to estimate the tendencyto recurrence after operation at a far lower figurethan the 50% of Gross.9FIBRO-EPITHELIAL GROUP — EPITHELIAL TYPE

—CYST-ADENOMA.

A certain number of tumors of the fibro-epi-thelial group present such marked epithelialchanges as to merit description under anotherheading separate from the tumors of the peri-ductal fibrous type. Tumors of this natureoccur in two main classes; both of which, how-ever, are much rarer than the tumors of thefibrous type. For these two classes I haveadopted the names (l)fibro-cyst-adenoma, and(2) papillary cyst-adenoma. The tendency tothe formation of cysts in all inflammatory or

neoplastic conditions of the mammary glandhas been the cause of great confusion in nomen-clature. A sharp line of differentiation, however,should be possible between a tumor of localunicentric growth and a diffuse process involvingmultiple areas in the gland or even the breastof the other side. For this reason I have seenfit to reject the term of " cyst-adenoma," for allbut local new growths of a cystic character, themajority of which are tumors which come underthe clinical classification, the papillary type.Multiple diffuse cyst formation in the breast Iprefer to class as abnormal involution, and underthat heading I will go more at length into thenomenclature of other writers. The occurrenceof cystic tumors among the periductal fibromatahas already been remarked, and allusion was theremade to their purely secondary character.

FIBRO-CYST-ADENOMA.

In four cases in this entire series secondary epi-thelial new growth occurred in a periductal fibroustumor, producing a combination of lobular andcystic epithelial growth in a troma of periductaltissue. To these tumors I have applied the nameof " fibro-cyst-adenoma." They may be regardedas an epithelial exaggeration of the periductal



fibroma and their anatomical structure differsfrom the fibroma only in the secondary prolifera-tion of their epithelium. Tumors of this char-acter have been described by other writers chieflyas adenomata, papillary cyst-adenomata, cysto-adenoma-proliferum, poly-cystoma, cystic fibro-ma, tubular-adenoma, etc. The characteristicsof the fibro-cyst-adenomata were as follows:The size of the tumors varied from the size of awalnut to that of a fist. They were lobular andof hard consistency, and movable under the skin.They had in general the appearance of a peri-ductal fibroma in which the epithelial lined slitsgreatly preponderate. The axillary glands werenot enlarged and the nipples were not affected.On section these tumors presented a definitecapsule and a lobular structure containing cystsof varying sizes. The cysts presented papillaryoutgrowths of connective tissue with a coveringof epithelium. Microscopic examination showedmarked proliferation in the epithelial cells liningthe cyst cavity and covering the papillary out-growths. The ducts appeared to be more in-volved in these tumors than the acini of the gland.All occurred in young single women, the averageage being thirty-two. They were tumors of slowgrowth and of long duration. Pain and dischargefrom the nipple were not recorded. All of thesetumors were removed by excision without ampu-tation of the breast.

The prognosis of tumors of the fibro-cyst-adenoma type must be regarded as entirelyfavorable, and removal of the tumor without am-

putation of the breast gives freedom from recur-rence. It must be said, however, that the indi-cation for operation in these tumors is even greaterthan in the fibromata, because of the epithelialproliferation which they present and the possi-bilities of its later development into carcinoma.

PAPILLARY CYST-ADENOMA.

The last group of tumors of the fibro-epithelialtype comes under the heading of " papillarycyst-adenoma." These tumors are perhaps themost distinctive in clinical characteristics of anyof the benign neoplasms of the breast. Ninecases occurred in the hospital series and three inprivate practice. They thus form the majorityof the tumors which go to make up the epithelialdivision of the fibro-epithelial group. Tumorsof this nature have been described by differentwriters under many headings. Gross 10 in hisclassification in 1880 attempted to restrict tothem the term " adenoma " or " true adenoma,"with the intention of indicating the growthof glandular epithelium in the form of tubesand ducts which they present. The Englishwriters u 12 13 describe these tumors under theterms villous papilloma, duct papilloma; or bywhat, in the light of modern investigation, wemust consider the misnomer " duct cancer."By the German writers u * 15 7 16 they are com-monly known as cyst-adenoma-papillare, poly-cystoma proliferum, intracanalicular cyst-ade-noma, proliferating cyst-adenoma, or papillarycystoma.

To my mind the term papillary cyst-adenomais the most satisfactory name for tumors of thisvariety. The histologie picture of a papillaryor warty outgrowth is that of a connective tissuepedicle surmounted by an epithelial covering.Papillary structures of this character are thedistinguishing feature of the tumors of this groupand serve to differentiate them absolutely fromother benign tumors of the breast. Tumors ofthis character rarely attain great size. Theirconsistency is hard although fluctuation mayoccasionally be detected. Adherence to theskin and enlargement of the axillary glands arenot to be expected. The situation of the tumorin the breast is generally beneath or in close rela-tion to the nipple. On gross examination thesetumors present the picture of a cyst cavity con-

taining papillary or villous outgrowths from itswall. The fluid contents of the cyst is generallyhemorrhagic. Microscopic section shows thecharacteristic papillary outgrowth of connectivetissue surmounted by a luxuriant growth ofepithelium. The epithelium in these cases pre-sents the characteristics of ductal rather thanacinal epithelial cells. The average age of thecases of tumors of this type was fifty-two years;and married women, especially those who haveborne children and reared large families, appearto be especially predisposed. As a rule, thetumors are of slow growth and of long duration.The most characteristic symptom, however, isthe discharge of bloody fluid from the nipple,which occurs in the majority of cases.

Amputation was done in three cases of papil-lary cyst-adenoma and resection and excision insix. The absence of recurrence after operationis strong evidence in favor of the right of thetumor to a place in the category of benign neo-

plasms.Papillary cyst-adenoma is thus seen to be a

tumor occurring late in the life history of thebreast, and one to which the functional activityof the gland in a certain measure predisposes.They are comparatively rare, but their signifi-cance must not be overlooked on that account,and it should be noted that, although only ninecases occurred during this decade in the hospitalseries, during that time two cases of cancer werefound in which this disease was also present. Infact papillary cyst-adenoma has been thoughtto predispose to a special form of cancer, " adeno-carcinoma-destruens " (Tietze 16), and this con-dition was present in both of the cases to whichI have referred. The prognosis then of thesetumors with early and complete removal is good.If not removed by operation, however, the dangerof malignant sequelae is very great.

HYPERPLASIA.

The term " hyperplasia " I have reserved inthis classification for those processes in the breastwhich are of a diffuse character and for this reasonto be sharply differentiated from the local uni-centric process of actual tumor formation.

By hyperplasia is meant an increase in thefibrous and epithelial elements of the breast,



affecting one or more lobules of the gland, andpresenting the microscopic picture of an increasein the number of cell elements of both of thesignificant tissues, the fibrous stroma, and theepithelium. Under this category two main divi-sions can be made: (1) "diffuse hypertrophy "

of the breast, and (2) abnormal involution.

DIFFUSE HYPERTROPHY.

The so-called " diffuse hypertrophy " of thebreast is a rare disease, but one which has receivedattention from many writers.17 Two instancesof this disease are included in the hospital seriesof cases, and fortunately the two cases serve toillustrate well the two main sub-varieties of thiscondition.

One form of diffuse hypertrophy, and perhapsthe most common variety, is that which occursin connection with pregnancy.

Of this form we have an example in the caseof a woman of forty-three, who had undergonetwo miscarriages and who presented the abnor-malities and diseases of the pelvic organs whichhave so frequently been noted in connectionwith this disease. Both breasts were affectedand were uniformly enlarged. The consistencyof the breasts was soft, but hard, stringy nodulescould be detected on palpation. Microscopicexamination showed the presence of fat, loosefibrous tissue and atrophied gland structures.There were no cysts nor discrete fibromata. Thebreasts gave trouble only from their size, and thepain was of an insignificant character; a doubleamputation was performed.

The other case serves to illustrate the conditionas it appears in younger women. This occurredin an unmarried girl of seventeen, and involvedthe left breast only, which was enlarged to twoor three times the size of the breast on the otherside. The consistency of the breast was nodularand corded, and tenderness or " mastodynia "

in this case was a marked feature. Irregularityof catamenia was noted, but no other abnormalityof the pelvic organs could be determined. Inthis case no operation was done and a year anda half later, under palliative treatment by sup-porting bandages, the enlargement of the breasthad partially subsided and tenderness and mas-

todynia had diminished to a moderate extent.

ABNORMAL INVOLUTION.

The hyperplastic changes in the tissue of themammary gland which accompany the processof involution have long been the subject of con-tention and controversy among pathologists andclinicians. The chief source of difficulty in theclassification of the involution changes of thebreast is undoubtedly to be attributed to thetendency which the gland exhibits to the forma-tion of multiple cyst cavities. In all probabilitya moderate degree of dilatation of the ductsmay be considered the normal condition of thebreast at a certain stage of its decline, and onlywhen these cysts and their surrounding stromabecome so extensive as to attract attention by

the induration which they produce does theprocess attain a degree of abnormality sufficientto classify it as pathologic. Multiple cysticdisease of the breast has long been recognized,and has been many times described under a mul-titude of different terms. Abernethy and Bellconsidered this a form of hydatid disease, as didalso Sir Astley Cooper. Brodie took the viewthat the cysts were caused by dilatation of thepre-existing ducts, and the French surgeons ofthat time agreed in this interpretation. Pagetbelieved that cysts were formed in the breastboth by obstruction and by active new growth,and the German school of pathologists of thattime were inclined to attribute the multiple cystsof the declining breast to an active new growthof the acinal eptihelium, although Virchow wasa believer in the obstructive origin of mammarycyst formation. A new impetus was given tothe study of this disease by Reclus,18 who in 1860attempted to show that the cysts were producedby an active new growth of the epithelium ofthe gland closely allied to carcinoma, and gaveto it the name of intra-acinous cystic epithelioma.Later publications in 1887 called attention morespecifically to the epithelial origin of these growthsand the disease became known in France as the" maladie de Reclus."

From the time of Reclus's later publicationsmany different theorets have been advanced toexplain the occurrence of multiple cysts in theinvoluting breast. These theories may begrouped into three main classes, which maybe briefly described as follows :

—1. Inflammatory Theory.—The prevailingtheory among French 19 20 21 and English sur-

geons,22 n 2i 25 and the view which has beenvigorously supported by Koenig24 and otherGerman surgeons,25 attributes the cyst formationto inflammatory origin, " chronic cyst mastitis."Whether this inflammatory process arises frominfection of the ducts, as believed by the Frenchwriters, or whether it is of a more remote char-acter, is not determined. The chief points on

which the theory is based are the evidences ofchronic inflammation in the round-cell infiltra-tion and thickened stroma of the gland, and inthe clinical course of the disease, which is thoughtto be subject to repeated exacerbations. Koeniglays great stress upon the increased swelling andsensitiveness at the time of the catamenia.

2. New-Growth Theory.—

The second theory,advanced originally by Schimmelbusch,7 and sup-ported by Sasse,15 Maly,26 and to a certain extentby Tietze,16 attributes the process to an actualnew growth. This theory is widely acceptedin Germany, and has obtained some support inother countries. Because of the supposed rela-tion to tumor formation, the names of cyst-ade-noma (Schimmelbusch7), poly-cystoma (Sasse15),or cystoma (Tietze16), have been proposed. Itshould be stated that Sasse and others attemptto distinguish the lesser degree of this disease,under the name of " involution cyst," from themore advanced process of " poly-cystoma."The line, however, between the two conditions



appears to be determined with the greatest diffi-culty.

3. Involution Theory.—

A later theory, andthe one which has been supported by the investi-gations which have been carried on in this coun-try, rejects both the inflammatory and the neo-

plastic origin, and refers the proliferation of thebreast epithelium to the abnormal conditionsproduced by involution. This is, in a measure,a return to the views of Reclus, who recognizedthe significance of the epithelial proliferation,but erred, in considering this the origin of theprocess.

Tietze 10 agrees that the disease is chiefly epi-thelial in character, and of a more diffuse dis-tribution than would warrant the title of a tumor.Bloodgood 27 has recognized this fact in the namewhich he suggests of "senile atypical parenchy-matous hypertrophy." Greenough and Hart-well 28 made a study of 30 cases of this diseasein 1902. By these writers the chronic appear-ances of the stroma of chronic cystic disease ofthe breast are considered to be of secondaryimportance, whereas the diffuse character of theepithelial proliferation is thought to be evidencesufficient to prevent the interpretation of theprocess as an actual new-growth. Accordingto their views, the connective tissue increase isattributed to the normal process of involutionand atrophy which the breast gland shares duringits declining stage, with other organs of the geni-tal apparatus. Cystic changes of the ducts andacini are produced by obstruction from this thick-ened fibrous tissue. Up to this point the processmay be considered physiologic. It has long beenunderstood, however, that the epithelium of thebreast gland is of notoriously unstable equilibriumand the abnormal condition of cyst formationprovides a stimulus which in about half the casesof this disease, as nearly as can be estimated,leads to secondary proliferation of the epithelium.Were we able to state at what point this epithe-lial proliferation would cease no danger fromthis disease need be apprehended, but it has beendetermined by a number of different observersthat in the proportion of about one case in tenthis epithelial proliferation oversteps the boundsof hyperplasia into the territory of malignantnew-growth. It is probably to these prolifera-tive cases of abnormal involution that Sasse andSchimmelbusch would restrict the term poly-cystoma or cyst-adenoma, and it must be con-fessed that a certain number of advanced casesof abnormal involution can be distinguishedwith great difficulty from actual new-growth.It has seemed to me, however, to be safer to re-strict the term " cyst-adenoma " to tumors of'local origin and well-determined outline, and toapply the general term of hyperplasia or abnormalinvolution to all of those diffuse processes whichby their microscopical appearances are still withinthe pale of benign diseases of the breast.

The chronic interstitial mastitis of the Englishwriters,22231112 chronic cirrhosing mastitis (Bill-roth 6), cystoma (Sasse15), or cystoma (Tietze 16)have been tic epithelioma (Reclus18), maladie de

Reclus,10 maladie noneuse de la mammelle (Pho-cas21), chronic cystic mastitis (Œnig2i), periductalmastitis (Delbert20), poly-cystoma epitheliale(Sasse15), cyst-adenoma (Schimmelbusch7),plexi-form fibroma (French19 writers), chronic mas-titis or diffuse fibro-adenoma (Wood29), cysticdisease of the breast (Bryant "), and fibrous andglandular hyperplasia with retention cysts,

—are but a few of the designations which havebeen applied to this disease. The last term,which was evolved by Dr. W. F. Whitney, patholo-gist to the Massachusetts General Hospital, isprobably the most correct and satisfactoryappellation for this condition which could bedevised, but suffers in being too unwieldly forpractical use. For this reason I prefer the nameof " abnormal involution."

Eighty-seven cases of abnormal involutionappear in the hospital series and 28 in the casesfrom private practice. Of this total number59 show lesions which are classified as cysticonly, the remaining 56 presenting evidence ofepithelial proliferation. A certain amount offibrous tissue increase with more or less round-cell infiltration is characteristic of involution ofthe breast under any circumstances.

The changes which have been classed as cysticare those presented by breasts in which thisfibrous thickening has produced dilatation ofthe ducts alone without obvious changes in thecharacter of the epithelium. The proliferativegroup are those which are characterized in addi-tion by proliferative changes in the epitheliumof the cysts or acini.

Of the proliferative group, three subdivisionshave been distinguished according to the char-acter or degree of the epithelial growth: (1)proliferation of the acini, (2) papillary outgrowdhsof epithelium into cysts, and (3) adenomatousproliferation of epithelium. The significanceof these three groups will appear in the discussionof their special characteristics. The gross ap-pearances of cases of abnormal involution maybe summarized as follows: A considerable pro-portion of the cases are diffuse in character,while others present local indurations of thebreast tissue of varying size and situation. Theconsistency of this induration is hard or nodular,although softer and even fluctuating areas are

occationally noted. It would seem, however,that a cyst to be appreciated with certainty byits elastic feeling or fluctuation must attainconsiderable size. Pain and tenderness are pres-ent in about half of the cases, and although notdiagnostic this symptom must be considered ofsignificance in view of the generally painlesscharacter of carcinoma. Axillary glands aredetected in only a small proportion of the cases.It has been my impression, based on clinicalexperience, that inversion of the nipple is a fre-quent accompaniment of abnormal involutionand it may certainly be stated that previousdiseases or abnormalities of the breast, such le-sions especially as interfere with the normalfunction of the gland, are predisposing factors inthe origin of this disease.



On gross examination these specimens showfibrous induration of the breast gland, with thepresence of multiple cystic cavities containingfluid varying from a clear serum, through all theshades of brown and green and black. The pro-cess is diffuse in the vast majority of cases, al-though one lobule of the gland is often noted as

the situation in which the disease is most ad-vanced. Under microscopic examination fibroushyperplasia and secondary involvement of thegland structures with cyst formation is observed.

The majority of cases occurred between fortyand fifty years of age. Married women were

apparently the more susceptible to the disease.The duration was a matter of considerable varia-tion, and the rate of growth was generally slow.Trauma was apparently of comparatively littleetiologic importance. In this series of cases therelation between exacerbations of the diseaseand the catamenial period was conspicuous by

.

its absence, only four cases having stated thatthe induration was influenced in its sensitivenessor in size by the monthly period. Dischargefrom the nipple was noted in 14 cases, and was

relatively more frequent among the prolifera-tive than among the purely cystic types of in-volution. The character of the discharge variedfrom yellow serum and watery or milky fluidsto brownish and bloody fluid and even pus. Ofthe 14 cases, however, which showed a dis-charge from the nipple, 7 presented tumors im-mediately beneath the nipple, and it is probablethat discharge from the nipple will be foundto depend more on the situation of the cyst thanon the character of the epithelial changes withinits wall. The right breast was involved in 27cases, the left in 42, and both breasts in 17. Thetendency to the involvement of both breasts inthis disease is worthy of remark. So far as thesituation of the induration in the breast tissueis concerned, the outer and upper hemispheresappear to be the seats of predilection; and theupper and outer quadrant, as in almost all dis-eases of the breast, is the region most commonlyaffected. Diffuse induration involving the wholebreast occurred in 14 cases. Operation was

performed in 86 of these 87 cases. In 41 theentire breast was amputated. In 45 an excisionor resection of the breast was done after themethod later to be detailed.

The subclassification of the cases of the pro-liferative type is of interest only from a histologiepoint of view, and can be determined only bymicroscopic examination.

The 36 hospital cases which presented prolifera-tion were subdivided as follows: Ten showedan apparent increase in the number of the glandacini, when compared with normal breasts ofthe same age, and conditions of marriage, child-birth, and lactation. This increase in the numberof acini was accompanied in the majority ofinstances by a proliferation of the epitheliumto such an extent that thickened or even solidcolumns of cells were produced, retaining, how-ever, the characteristic formation of the glandducts, and presenting no infiltration beyond the

basement membrane. The significance of¿thisform of proliferation could not be determinedfrom the cases which were available for exami-nation. But it is not improbable that it mayultimately lead to carcinoma. In support ofthis assertion one case of scirrhus cancer occurredin the hospital series and two in private practice,in which abnormal involution of the acinal typewas also present. The clinical characteristicsof the cases of acinal proliferation were not tobe distinguished from those of other types ofabnormal involution.

Another class of epithelial proliferation inabnormal involution is to be described underthe term of papillary proliferation. This groupagain is one of only histologie significance. Bythis term is meant a growth of epithelium incyst cavities of such a nature that the epithelialcells are heaped up and project into the cavity,without a connective tissue pedicle or support.The microscopic picture of this papillary projec-tion suggests strongly that more epithelium isproduced in the lining of the cyst wall than canbe accommodated on the basement membrane,with the result that certain portions are forcedout toward the center of the cavity. Therewere 18 cases of abnormal involution which pre-sented this form of proliferation. Of this classagain, the clinical characteristics are not to bedistinguished from those of other forms of abnor-mal involution.

The third and last type of epithelial prolifera-tion which is to be distinguished is that to whichTietze16 and Greenough and Hartwell28 haveapplied to the name of " adenomatous " prolif-eration. Adenomatous proliferation occurs onlyin cases in which papillary outgrowths are alreadypresent, and must be regarded as a somewhatmore advanced type of epithelial growth. Inthese cases the papillary outgrowths from differentportions of the cyst wall appear to meet and fusetogether, and the resulting microscopic pictureis that of a space filled with epithelial cells inwhich here and there an open gland lumen isobserved. This form of proliferation occurredin eight of the hospital cases and in four of thecases from private practice.

Adenomatous proliferation derives its greatestinterest, however, from the fact that it is chieflyin its presence that we find the combination ofinvolution and carcinoma. During the ten yearsin which the 12 cases of adenomatous prolifera-tion have been collected, 517 cases of cancer

have come under observation. Of this numberthere were 15 which presented the lesions ofabnormal involution and carcinoma. Three ofthese cases showed scirrhus carcinoma in con-nection with abnormal involution of the acinaltype of proliferation. These have already beenconsidered. Of the remaining 12 cases, 9 showedadenocarcinoma with involution changes of thepapillary and adenomatous character. In thethree other cases medullary carcinoma of a moreadvanced type was recorded in the pathologicreport. Thus during the collection of 115 casesof abnormal involution 15 cases of this disease



in combination with carcinoma were obtained,or 13%. This figure is somewhat greater thanthe 10% which has been found by Tietze,16 Green-ough and Hartwell,15 and others, but may beconsidered as very closely approximating theactual condition.

Attention must again be called to the fact thatthe clinical symptoms of these proliferative typesof abnormal involution are to be distinguishedin no way from those of simple abnormal involu-tion, and by no sign on physical examination canthe degree of epithelial growth be determined.Microscopic examination is, therefore, absolutelynecessary for the correct classification of theindividual case, and in fact the variation betweenthe different portions of the same gland is sogreat that microscopic examination of all thetissues is necessary to a correct diagnosis.30 Forthese reasons and because of the liability to car-cinoma which these cases present, I have no hesi-tation in recommending operative treatment forabnormal involution in every case. An indura-tion in the breast gland at or about the timeof the menopause which is palpable with the flatof the hand against the chest wall I considersufficient indication for operation.

For cases of this character I perform " plasticresection of the mammary gland," by means ofwhich a thorough exploration of the gland maybe obtained and portions of a suspicious naturemay be removed for microscopic investigationwithout the mutilation of an amputation. Thisoperation I would recommend for every case ofabnormal involution in which the diagnosis canbe established.

CHRONIC INFLAMMATORY DISEASES.

The chronic inflammatory diseases of thebreast are divisible into two main classes, thoseoccurring in connection with lactation, and thoseproduced by specific inflammatory diseases suchas tuberculosis. It is to these diseases of thebreast that the much-abused term " chronicmastitis " should really be applied, but this termhas been so loosely employed that its use in anyway is at present inadvisable. In addition tothe two main classes detailed above, I haveventured to place in the category of chronicinflammation two cases of non-malignant eczemaof the nipple, and five cases of single retentioncyst of mechanic origin. It is to be regrettedthat no case of syphilis or actino-mycosis enteredthe hospital during this decade, as it is underthis heading that these diseases should be classi-fied.

There were eight cases of chronic lactation mas-titis and 13 of tuberculosis in the hospital series.Chronic lactation mastitis has been again sub-divided into five cases of chronic abscess andthree of inflammation of the ducts (ductalmastitis). The cases of ductal mastitis showmultiple small abscesses and necrotic foci dis-tributed through the breast gland in close re-lation with the larger ducts. Clinically theypresent tender indurated masses in the breast.The mass may be adherent to the skin, and

enlarged axillary glands may be present. Micro-scopically these breasts show an increased amountof fibrous tissue with small yellowish foci of necro-sis and abscess cavities, while the tissue aboutthe ducts is infiltrated with great numbers ofsmall round cells and leucocytes. The diseaseespecially affects young women, and may beattributed to infection of the ducts during theprocess of lactation. The chief clinical char-acteristics are diffuse induration of the breastoccurring shortly after lactation, with a ten-dency to involvement of the skin and more painand tenderness than would be expected with atumor. In two of our cases the breast wasamputated, and in one the affected lobule wasexcised.

CHRONIC ABSCESS.

Five cases of chronic or subacute abscess ofthe breast occurred during the decade. Allpresented a tumor or nodular induration of thebreast of varying size, and in only one case wasfluctuation to be detected. Tenderness, adher-ence to the skin and enlargement of the axillaryglands were occasionally noted. The patho-logic examination showed one or more abscesscavities of varying size, and microscopic exam-ination showed excessive infiltration of the glandtissue with leucocytes and round cells. Thepatients were all young adults. All were married,and four had had children, while one had neverbeen able to nurse her child. Two gave a his-tory of previous acute breast abscess. The ab-scesses were, as a rule, of slow development, andpain was not a pronounced symptom. In twocases old discharging sinuses from previous acuteabscesses were still present. In three casesincisions with drainage were made; in one casethe breast was amputated, and in one the affectedlobule was excised.

GALACTOCELE.

One of the cases of chronic abscess presented,in addition to the abscess, the picture of a reten-tion cyst in a breast which was in active lacta-tion. This condition has been repeatedly de-scribed under the name of galactocele, but itsoccurrence is comparatively infrequent, as maybe seen from the fact that but one instance oc-curred in our total of 758 cases. The symptomsof galactocele do not differ appreciably fromthose of single retention cysts, except that thecondition occurs in a breast which is yet, or hasbeen recently, in active lactation. Under thesecircumstances the contents of the cyst is of a

milky rather than of a serous character, becauseof the milk-secreting function of the cells whichline the cyst at this period of the breast's devel-opment.

TUBERCULOSIS.

In this division are included the thirteen casesof chronic disease of the breast produced by thetuberclebacilli. Indurated masses of varyingsize were produced in the breast, of irregularoutline and varying consistency. Tendernessand fluctuation were present in most of the cases.



The tendency to involvement of the skin wasmarked, and in seven cases there were dischargingsinuses present at the time of operation. En-largement of the axillary glands was generallynoted. Tuberculosis of other organs was presentin only three cases. Microscopic examinationof the specimens showed the characteristic tuber-cular tissue with cheesy foci of necrosis and thetypical histologie peculiarities of tuberculosis.

The average age was thirty years, and in threecases the patients were under twenty years ofage, although the disease was well distributedthrough the decades of adult life. Marriedwomen who had borne children appeared to bethe most susceptible, and in two cases acuteabscess had occurred during a previous lactation.The average duration was nine months and thedevelopment of the disease was of a chroniccharacter. Pain was absent in about half thecases. Both breasts were involved in only one

case; and the upper and outer quadrant was againthe seat of predilection. In seven cases thebreast was amputated, in four a portion was

excised, and in two the abscess was simply openedand drained. The immediate result in all ofthese cases was satisfactory.

SINGLE RETENTION CYST.

Five cases occurred in the hospital series whichpresented the characteristics of a single cyst, ofvarying size, with no abnormality of the sur-

rounding tissues. These cases have beenseparated from those of abnormal involutionbecause of their purely local character and theabsence of pathologic changes in the breasttissue in their immediate vicinity. It is not tobe denied that a certain number may belongproperly among those of abnormal involution,and that a more extensive examination mighthave given data which would justify their inclu-sion among the latter class. In the absence ofthese data, however, it has seemed wiser to placethe single retention cysts in a separate class.Cysts of this character have been described underthe name of " evolution " cysts by Gross.10 Theydiffer in no way from the simple cysts producedin other glandular structures by accidentalocclusion of a duct. We are especially familiarwith cysts of this type in the epididymis and thesalivary glands. Occlusion of a duct may bedue to any one of a number of different causes;congenital defects, trauma, and the scars oftransient inflammatory processes of the nippleor in the substance of the gland, being.perhapsmost frequently observed. Operations on thebreast for small fibromata are not infrequentlyfollowed by cicatrical obstruction of the ducts.

The five cases here presented varied fromtwenty-one to forty-nine years of age — threewere single and two married women. The dura-tion was comparatively short (an average of abouttwo months) and development was rapid; therewas no pain or tenderness as a rule, and theclinical appearances were those of a single, pain-less, elastic, fluctuating tumor, varying in sizefrom a walnut to a hen's egg, and of uniform and

rapid growth. The axillary glands were notenlarged. Excision was done in every case, anda single thin-walled cyst removed without otherabnormality of the gland. The fluid containedin these cysts varied from clear serum to a fluidof yellowish-brown color

—

with more or less tur-bidity. The significance of a single retentioncyst in the breast must be considered very slight,and their chief interest and importance lie inthe difficulty in distinguishing them from mul-tiple cysts of abnormal involution. Because ofthis difficulty, exploration and excision are tobe advised.

NON-INDIGENOUS TUMORS OF THE BREAST.

A certain number of tumors of the breast occurwhich are situated in that organ only acciden-tally, as it were, and have no special relation tothe mammary gland as regards their origin.Among these tumors I have classed the sarco-

mata, which were derived not from the periductaltissue but from other sources, and which pre-sented no evidence of inclusion of the glandularelements of the breast. Two of these tumorswere classified as fibro-sacroma, one was ofmixed cells and one of spindle cells. All weremuch smaller in size than the periductal sarco-mata, and all appeared to involve the breastgland before removal. The tendency to theinvolvement of the skin in these tumors was

marked, two being actually ulcerated and oneother showing marked adherence and brownishdiscoloration. The ages varied widely fromthirty-three to seventy years of age. All werein married women. The duration varied fromsix months to ten years. Two were of slowgrowth and two more rapid within the last fewmonths. Discharge from the nipple was notnoted and the axillary glands were found enlargedin only one case. Pain was generally present.Amputation was done in three cases and exten-sive excision in the fourth. The origin of thesetumors appeared to be from the connective tissuesforming the stroma of the mammary gland.The essential anatomic elements of the mam-mary gland were merely displaced by their growthand they differ in no respect from sarcoma of asimilar character arising in other organs. Theirsize was smaller than the periductal sarcoma,but with this exception the two forms were withdifficulty to be distinguished.

The other non-indigenous tumors of the breastwere lipoma and lymphangioma. Three casesof lipoma occurred in the hospital series and onein private practice. They were freely movableand not tender. No enlargement of the glandsor abnormality of the nipple could be detected.They were multiple in one case. The ages ofthe patient varied from twenty to sixty-six years.The duration of the tumors from three to twentyyears. They were of slow growth and occasionedno objective symptoms. All of the cases were

operated on, excision being done in two and am-

putation in a third.A single case of lymphangioma occurred during

this decade. This was a cavernous lymphan-



gioma in a girl of sixteen, which started in theaxilla and involved the breast only by secondaryextension. It was of six years' duration. Ex-amination showed a dense fluctuating tumorthe size of a man's head occupying the positionof the breast, but being situated behind the glandand extending from there to the axilla and intothe neck. The whole mass was removed withoutamputation of the breast and the patient madegood recovery.

SUPERNUMERARY BREAST.

One instance of this condition is included in theseries of cases from private practice, and anotherwas accidentally discovered in one of the hospitalcases of periductal fibroma. Both occupied asituation at the anterior margin of the axilla;and this appears to be the most common site,although supernumerary breasts in other situa-tions along the so-called " milk-line " from theaxilla to the groin, are occasionally reported(Young31).

Both cases presented a small collection ofglandular tissue without a nipple. No symptomswere produced in either case, and in one the tumorwas discovered only in routine physical examina-tion. In the other anxiety was aroused by thepresence of the axillary tumor, but was promptlyrelieved when the correct diagnosis was ascer-tained.

In neither of these cases was there any evi-dence in support of the theory that supernume-rary breasts are more prone to tumor formationthan breasts in a normal condition; and indeedthis theory is probably one of those which havebeen handed down from writer to writer withoutsatisfactory evidence for its support (Martin32).The rarity of cases of supernumerary breast(only two occurring in 758 cases) is the probableexplanation of this uncertainty in the matterof prognosis.

PLASTIC RESECTION OF THE MAMMARY GLAND.

As Stewart M justly says, there is a lamentableand unnecessary difference of opinion as to thetreatment of doubtful tumors of the breast.

Many of the tumors are situated in such portionof the gland as to involve disfiguring scars in a

conspicuous locality if an incision is made directlyover the tumor. Many operations designedfor benign tumors involve an amputation of thebreast. This unfortunate organ has sufferedcountless mutilations in times past in conformitywith prevailing customs. Thomas 24 first soughtto overcome this difficulty by proposing an inci-sion along the lower and outer border of thebreast, but details as to his technic are wantingin the original article, and the method does notseem to have been generally adopted.

In the treatment of cysts Abbé 35 has advocatedan exploratory puncture, and gives a large num-ber of permanent cures effected by simply evacu-

ating the cyst. McGraw36 has also written uponthe treatment of cysts by aspiration, and in themain agrees with Abbé, although he notes twocases in which he observed the development of

cancer after this procedure. Bull37 also haslong been an advocate of aspiration of cysts, andhad no cases of cancer following aspiration inhis own experience. McGraw's two cases, how-ever, are excellent concrete examples of thetheoretical dangers of aspiration.

At one time I was in the habit of using explo-ratory puncture to determine the nature of adoubtful growth, and for this purpose used aMixter punch (a long steel cylinder of smallcaliber with a handle at one end and sharp edgesat the other). Richardson, however, later re-ported cases in which an extension of the can-cerous process. was observed along the line ofpuncture; and although I have succeeded incuring many a cyst by puncture, I have aban-doned it as a routine practice. In doubtfulcases it may cause serious displacement of cancer

cells; in cases of cystic disease it disposes onlyof the larger cysts. I should still use the method,however, in exceptional cases, such as in thepresence of a large cyst near the nipple in a smallbreast which could be easily palpated.

Herbert SnowM has recently recommendedremoval of cystic tumors of the breast by forciblemassage. Cysts can undoubtedly be rupturedin this way and permanently cured, but it is nota method of precision, and would be most dan-gerous advice for general adoption.

StewartM advises an exploratory procedureso planned as to avoid contact of cut lymphaticswith any tissues left behind, and any chance ofdissemination of cancer cells by pressure beforethe removal of the breast.

Bloodgood27 agrees in advising exploratoryincision, and states that this operation is theroutine treatment of doubtful tumors at theJohns Hopkins Hospital. Curtis 29 and Morris w

support this recommendation and Morris callsattention to the case with which consent maybe obtained for an exploratory operation earlyin the course of the disease, when the possi-bility of avoidance of a total amputation can beurged.40

The operation which I have gradually elabo-rated, starting with the so-called Thomas incision,consists in the removal of a V-shaped portion ofthe gland containing the cyst or tumor, andexploratory radiating incisions in the remainderof the gland. In the early operations I under-took to remove all sections of the gland involvedin cystic changes, and in several cases this resultedin removal either of the entire gland or of all butfragments here and there of the cortical portion.These fragments were brought together withcatgut sutures placed at certain points, or by a

single purse-string suture, so that in the breastsof women possessed of a fair amount of adiposetissue the organ could be reconstructed and thenipple (which previously perhaps had been in-verted) could be restored to its normal shape.In this way the outward shape and appearanceof the breast could be maintained while the entiremammary gland had been removed. Furtherexperience, however, showed me that this radicalmeasure was in the great majority of cases un-



necessary. The details of the operation as I nowperform it are as follows:

—The preliminary incision is longer than thatcontemplated by Thomas, and, starting at thelower border of the breast, opposite the middleof the outer arc of the lower inner quadrant,it runs along the lower fold and outer margin tothe inner border of the axilla, thus severing thelymphatic connections of the breast with theaxillary plexus of lymph glands. The incisionshould be carried down to the lower border of thepectoralis major muscle, which should be freelyexposed. The dissection is then carried alongthrough the loose connective tissue which liesbetween the pectoral fascia and the posteriorlayer of the fascia in which the mammary glandlies. With the left hand the operator reflectsthe breast upward and inward so that the pos-terior surface of the gland becomes exposed inits entire length. The gland tissue can now beseen through the transparent fascia and easilyinspected. If cysts are present they are readilyobserved, as the majority of them lie in the pos-terior portions of the gland. Usually there areone or two large cysts lying in the same quadrant.This portion can be removed by a V-shaped inci-sion without opening the cysts. The apex ofthe V lies directly under the nipple in the centralportion of the gland. Radiating from this pointincisions can be carried into the gland tissue inall directions, exposing and bisecting all smallcysts, so that no cyst, however small, remainsthat has not been laid open by the knife. Asecond V incision may occasionally be necessary,but this is rarely the case. The next step of theoperation, after arresting hemorrhage from thelarger vessels, is to close the V-shaped openingwith two rows of catgut sutures, one along itsanterior borders, and one bringing its posterioredges into contact. The gland is now releasedfrom the hand of the operator and dropped backon to the pectoral muscle, and it will be foundthat the various incised portions of the glandresume their natural positions and fit accuratelytogether. A few sutures may be needed to arresthemorrhage from small arteries in the gland sub-stance, but this is not usually necessary. Thegland should next be anchored to the subjacentpectoral muscle at its outer edge, and still anotherrow of sutures is advisable to hold together thedeep layers of the superficial fascia before closingthe outer edges of the wound with silkworm gut.The buried sutures are useful in removing strainfrom the surface suture.

The dressing should be applied so as to pro-duce lateral compression of the two hemispheres,as the ordinary swathe tends to tear the buriedgland sutures apart. I have devised for thispurpose a bandage which crosses in front and ispinned like a diaper; the ends of each half arethen caught into two loops which are attachedto suspenders crossing over the shoulders (empirebandage). This bandage can be elaborated intoa breast supporter, which is made for me byMessrs. Leach & Greene of Boston. If thebreast operated on contains a benign tumor, the

tumor can be removed by the V-shaped incision,the double suturing of the gland incision beingcarried out as before. I have found it betterto make a clean V cut even in the cases of theperiductal fibromata, as I have always found thecapsule of these tumors more closely adherent tothe gland tissue than is usually represented tobe the case.

In case of exploratory operations for cancerthe incision can be carried a little farther fromthe gland border if necessary and sufficientlydeep to sever all lymphatic connections with theaxillary lymphatics or those which run throughthe pectoral muscle. The incision into the pos-terior surface of the gland, if cancer be found,should be closed by a single suture before pro-ceeding with the major operation.

I have called this operation plastic resectionof the mammary gland, as the incision and sub-sequent sutures are devised with special referenceto restoring the gland as nearly as possible to itsnormal shape and to improve often the outwardappearance of the nipple. In case of inversionof the nipple, as is often seen in abnormal involu-tion, the gland tissue should be so laid open frombehind that a subcutaneous purse-string suturecan be placed around the base of the nipple andforce it outward.

I have performed plastic resection of the glandin 66 cases, in 8 of which both breasts were oper-ated on, so that I have performed the operationin all 74 times. The great majority of theseoperations were for abnormal involution, butin no case have I had any recurrence of cystformation. It is remarkable how completelyall trace of the incisions into the gland tissuedisappears. It is usually difficult to tell fromthe feel, only a few months after the operation,what portion of the gland has been resected. Asthe outward cicatrix is practically invisible tothe patient, the result from a cosmetic point ofview is also most satisfactory. In three cases

only have I found suppuration of the gland totake place, and this was attributed to infectionfrom the cavity of the cyst. One was a case oflarge multiple cysts, another was a large cystnear the nipple, and the third was a case of papil-lary cyst-adenoma near the nipple in an oldwoman. I have tested many exposed cysts forbacteria, but have invariably obtained negativeresults. I should, however, advise the operatorto avoid opening the larger cysts if possible, andto use catgut for deep ligatures and sutures, as

they are quickly absorbed and leave no trace oftheir presence in the sensitive structure of thegland. I have had stitch abscess in the skinsutures in two cases only, neither of which wasin private practice.

As will be seen from my experience in 74 cases,the operation is attended with little risk or dis-comfort to the patient, who may be assured thatit can be performed without a prolonged con-valescence and without disfigurement of anykind. I venture, therefore, to offer it to theprofession as a satisfactory substitute for thedisfiguring exploratory incision on the anterior



surface of the breast, for the uncertainties ofpuncture, and for the mutilation caused byamputation.

REFERENCES.

1 Collected and analyzed by Drs. R. B. Greenough and C. C. Sim-mons, whose valuable assistance in the preparation of this classifi-cation I wish here to acknowledge.

2 Warren : International Textbook of Surgery, second edition,vol. ii.

3 Warren: American Textbook of Pathology. Pathology of the

4Ribbert H.: Allgemeine Pathologie, Leipzig, 1901. F. C.W.Vogel.

ñ Ehrhadt: Deut. Zeit. f. Chir., December, 1899, vol. liv, p. 130.6 Billroth: Deutsche Chirurgie, vol. xli, Krankheiten der Brust-

drüse.7 Schimmelbusch: Archiv für klin. Chirurgie, 1892, vol. xliv, pp.

117 and 102.8 Schimmelbusch: Deutsche Gesellschaft für Chirurgie, 1890, p.

117.M Gross, S. W.: American Jour. Med. Sei., July, 1887.10 Gross, S. W.: Tumors of the Mammary Gland, New York, 1880.11 Bryant: " Diseases of the Breast," Cassel & Co. London, 1887.,2Sheild: " Diseases of the Breast," London. Macmillan & Co.,

1898.13 Williams: " Diseases of the Breast," London, 1894. John Bale

& Sons.14 v. Angerer: von Bergmann'«, Bruns' and Mikulicz'Handbuch

der Prakt. Chirurgie, vol. ii, p. 569.15 Sasse: Archiv, für klin. Chirurgie, 1897, vol. liv, p. 1.16 Tietze: Deutsche Zeitschrift für Chirurgie, 1900, vol. lvi, p. 512.17 Kirscheim: Archiv, für klin Chirurgie 1902, vol. lxvii, 2, p. 582.18 Reclus: Gazette des Hôpitaux, July 7, 1887, p. 673.19 Binaud and Braquehaye: Dentu and Delbet's Traité de Chir-

urgie, Paris, 1899, vol. vii.a) Delbet: Bull- de Soc.^d'Anat. de Paris, January, 1893, Paris.21Phocas: Report of Fourteenth French Congress of Surgeons,

Paris, 1901, p. 470.22 Banks: Lancet, 1900-1, p. 843, and 1902-1901, p. 309.23 Paul: Transact. London Pathological Soc, 1901, vol. iii, p. 30.24 Koenig, F.: Lehrbuch der SpecieUen Chirurgie. (Sixth edition),

vol. ii, Hirschwald, 1893, Berlin; Centralblatt für Chirurgie, 1893,vol. xx, p. 49.

25 Roloff: Deutsche Zeitschrift für Chirurgie, 1899, vol. liv, p. 106.^Maly: Zeitschrift für Heilkunde, 1898, xix, p. 355.27 Bloodgood: Jour. Amer. Med. Assoc, Aug. 6, 1904, p. 367.28 Greenough & Hartwell: Journal Medical Research, 1903, vol.

ix, No. 4, p. 146.29 Curtis & Wood: Medical News, New York, Aug. 13, 1904,

p. 294.30 The prognosis of a given case of abnormal involution untreated

is a matter of some uncertainty. Spontaneous rupture or dischargeof the cyst through the nipple may rarely occur, and the entire pro-cess subside for a time, or indeed, forever. Suppuration of the cystis occasionally observed. More often the disease becomes stationaryat a certain stage of its development or advances slowly involvingmore and more of the gland and the breast of the opposite side.Carcinoma develops

—

something more than 10% of all cases.31 Young, E. B.: Boston Medical and Surgical Journal, vol.

cl-, p. 319, 1904.32 Martin, E.: Archiv, für klin., Chirurgie, 1893, vol. xlv, p. 880.33 Stewart, J. Clark: The Journal, A. M. A., Aug. 6, 1904, p. 365.34 Thomas: New York Medical Journal, 1882.» Abbé: Medical Record, New York, Aug. 15, 1903.^McGraw: American Medicine, July 25, 1903, p. 142.37 Bull: Medical Record, New York, 1899, No. 16, p. 557.38 Snow: British Medical Journal, Oct. 17, 1903.39 Morris: The Journal A. M. A., Aug. 6, 1904, p. 368.40 Other references which may be consulted on the general sub-

ject are as follows: Clopton: Interstate Medical Journal* June,1904. Dreyfus: Virchow's Archiv., vol. cxiii, p. 535. Ellis: Annalsof Surgery, September, 1903. Renton: British Medical Journal,April, 1903. Warren: Annals of Surgery, December, 1904.

Original Articles.NOTES ON FRACTURES AND THEIR

TREATMENT.*BY F. J. COTTON, M.D.,

Assistant Surgeon, Boston City Hospital.I wish to bring to your notice not anything

radically new, but a number of matters in regardto treatment of fractures in which the point ofview has been changing the last few years tocorrespond to our larger knowledge.

The use of the x-ray has been very general forsome years and has not only taught us directlymuch about diagnosis and treatment, but hasalso greatly stimulated our observation of frac-tures in general and our attention to them.

The greatest direct service of the x-ray in myopinion has been that it has given us a cleareridea of the lesions in ordinary fractures, and hascleared up the pathology of various less usual frac-tures that previously were obscure. Of late therehas been a good deal said as to the routine use ofthe x-ray, some surgeons even going so far as totalk of malpractice in not using x-ray as a routine.This is sheer nonsense, of course, for it is neitherpracticable nor at all necessary to use x-rays inthe majority of cases in order to gain all diag-nostic data that will be of any use in treatment.Our treatment does not call for knowledge ofexact direction of fracture planes, etc., nor willit be changed by data regarding them. Beforex-rays were discovered very decent and satis-factory fracture work was done, and will stillbe done, so far as the usual run of cases is con-cerned.

It is, however, of the greatest service to usthat study of large series of x-rays have given usdata as to the type lesions in many fractures,such as those of shoulder, wrist, elbow, hip andankle. We are now in position to utilize pre-vious training in manual examination to muchbetter purpose. It is worth noting that in thehospital at least, dependence on x-ray examina-tion often results in superficial examination, a

tendency peculiarly bad in its effects on thetraining of house officers who now learn less thanthey should about diagnosis by other means thanthe x-ray.

There are type fractures (of the carpal scaphoid,in some cases of fracture of the tarsus and of themetatarsals, etc.) where x-rays are practicallyessential to all of us for certain diagnosis, to saynothing of the considerable number of cases ofsubperiosteal fractures, of bone cracks, etc.,that we can never prove except by skiagraph.

Apart from these special classes it is, I believe,our duty to examine and treat fractures with theaid of the x-ray if it is handy, without it in thevast majority of cases, unless as a later source ofinformation. The time when x-rays should beof most value is not before reduction, but after;not to show what we are dealing with, but whetheror no we have accomplished what we think wehave.

In all cases where there is any doubt as toproper position we should examine for positionwith great care, sometime about two weeks afterthe injury, at a time when errors can still readilybe corrected. At this time, unfortunately, thereis apt to be a good deal of swelling, often veryresistant swelling, and examination by the fingersis very difficult. This is where the skiagraph cando us the best of service, and its use should be cul-tivated far more than it is to-day, to review results,to determine whether gradual modelling, or re-

duction, or even open operation is called for.As a method of determining end results the

x-ray is, as has so often been said, very unsatis-factory. Despite all we now know academicallyabout distortions in the pictures they are practi-cally confusing to the most experienced at times— the picture may not represent the facts. Even*Read before the Norfolk South District Med. Soc., Feb. 2, 1905.



Documents

The Surgeon and the Pathologist