<<

THE AMERICAN JOURNAL OF CANCER A Continuation of The Journal of Cancer Research

VOLUMEXVI MAY, 1932 NUMBER3

OVARIAN NEOPLASMS

W. BLAIR BELL AND M. M. DATNOW

In the first part of this communication l some points in the of ovarian neoplasms were discussed. Here we shall first consider certain aspects of the clinical features associated with them. These cover so large a range of connected phenomena and conditions-namely, the symptoms and physical signs in a variety of circumstances which are related to the size and position, the complications, and the biological nature of the tumour concerned- that it will be neither possible, nor indeed desirable, to present a complete study in this place. Afterwards we shall examine some general principles in regard to the treatment of these neoplasms, especially in relation to the pathological and clinical features pre- sented, and to the age and condition of the patient. I1 CLINICAL FEATURES It is somewhat difficult to collate statistical information on a scale large enough to enable us to draw definite conclusions, even in respect of the average age and parity of the patients affected, for such figures do not seem always to have interested the collectore of

1 This paper, which is a continuation of that published in an earlier number of this JOURNAL(16: 1, 1932), likewise contains the substance (amplified in regard to statistics by M. M. Datnow) of the Introduction by W. Blair Bell to the Discussion on the subject at the British Congress of and held in Glasgow on April 21, 22 and 23, 1931, for an account of which see Journal of Obstetrics and Gynaecology of the British Empire, 38: 279, 1931. 439 440 W. BLAIR BELL AND M. M. DATNOW small numbers of cases. Yet it is only by putting together the findings in various series of small groups that we can obtain suffi- cient material for more comprehensive studies. We have made an attempt to collect from the literature and to tabulate certain features, and the figures given include analyses of some of our own material. We have not been through old hospital records, to which we have access, owing to the irregular manner in which the notes used to be kept, for, although certain figures relating to age would be obtainable, the pathological records are rarely adequate. Table VI relating to malignant tumours gives an analysis of forty-nine of our own recent cases and one museum specimen, of which we have full particulars.

PREVIOUSHISTORY For the most part the previous history, unless there be some- thing in it which is outstanding, is rarely recorded. Yet there are many points which would be illuminating had we a sufficient num- ber of cases on which to base our conclusions. For example, in the literature it is rarely possible to discover in the case of primary malignant cystic tumours whether the patient has had the tumour for many years. Among our own patients we have observed in- stances in which tumours had been known to exist for a number of years, and which on removal showed commencing and localized malignant changes in an otherwise innocent neoplasm. Since the supervention of carcinoma in an innocent neoplasm has been re- cently denied (44, 51), the question becomes of considerable im- portance both in respect of diagnosis and treatment. Among 232 cases of cystadenoma investigated by Smith (SO), Fleming (48), and ourselves, in 28-that is, 12 per cent-malignant changes had commenced. In many of these cases-especially in those of cystadenoma- in which changes are stated to have occurred, there is evidence of a more or less symptomless preexisting tumour for many years. In most a history can also be obtained of pain, moderate in severity, which is first occasional and at last almost constant, in the previous painless tumour. If relief be not given, the patient begins to lose weight and usually ascitic fluid accumulates. All these facts can be gleaned from the properly recorded histories of such cases. An example of a case of this kind is the following: S. B., aet. thirty-six years, was a widow and nulliparous. Menstrua- tion was regular 4/28. She had complained of swelling of the abdomen OVARIAN NEOPLASMS 441 for several years. Latterly there had been pain and loss of weight. There was a considerable quantity of ascitic fluid. A diagnosis of bi- lateral malignant disease of the was made. Spinal anaesthesia was administered and panhysterectomy with bilateral salpingoophorec- tomy was performed (Fig. 57). The patient made a good recovery and has remained well for t'wo years.

FI~.57. S. B., AET. THIRTY-SIXYEARS: BILATERAL PSIUDOMUCINOUS CYSTADE- NOMA, IN WHICHTHERE ARE MALIGNANTCHANGES IN THE SOLIDPARTS

Histological Examination: For the most part the tumours are innocent in the papillary portion, but here and there evidence is obtained, owing to the manner of epithelial proliferation, that the tumours are becoming malignant. Sections illustrating the earliest changes that occur when the biological nature of an ovarian neoplasm is altering have already been given (Figs. 42 and 46).

Innocent Tumours Cystadenomata: In records of 468 cases in which statements regarding age are made, we find that the youngest patient was seventeen years of age and the oldest was seventy-three. In our own cases the average age was 44.5 years; in Miss Fleming's, 42.4 years. Several other authors state that the commonest age is between forty-five and sixty-five. This, however, seems to be above the average, which from Miss Fleming's and our figures com- bined is 43.2 years. 442 W. BLAIR BELL AND M. M. DATNOW

Papillary Tumours: Miss Fleming, in a small group, gives the average age as 43.6 years. Erdmann and Spaulding (47), from an analysis of 44 examples give forty-one to fifty as the decade in which the largest number of cases is seen. Cystic Teratomata: In our patients the average age was thirty- seven years, while Miss Fleming gives an average age of 34.5. Smith states that 63.3 per cent of all cases occur between the ages of twenty-four and forty years, a wide distribution period. Fibromata: In our patients the average age was forty-four years. Miss Fleming gives 42.5 years as the average age. Smith states that 58 per cent of these tumours occur in women between the ages of thirty and fifty years.

Malignant I'umours Carcinorr~ata(primary and secondary): The average age of our patients was 40.5 years. Miss Fleming's patients were of an aver- age age of 47.1 years. In McIntyre's (49) series the average was forty-four years. In Bride's (45) patients the average was fifty- two years; in Whitehouse's (52) forty-seven years. Smith states that in half of all cases of malignant disease of the ovaries the patients are between the ages of forty and fifty-five years. Sarcomata: Williamson and Barris (53) state that "the greatest number of cases are met with about the age of puberty; the in- cidence is less during sexual life but rises again after the meno- pause." No figures are given. It is difficult to obtain statistics of value, for the lesion is comparatively rare and most observers include lepidomata (endotheliomata) and not tumours arising from connective tissues only. In our own small number of cases the average age was forty-six years.

These data bring to light information which should prove of considerable interest if it is verified by studies of larger numbers of cases in the future, and this doubtless will be possible with the better methods of recording now generally practised. The figures here given show that innocent ovarian tumours oc- cur at a higher average age than has usually been thought. Ex- cept in the case of cystic teratomata (dermoids), which would be expected at an earlier age in view of their pathology, the average age of primarily innocent tumours is definitely over forty years, and this average lies in the immediate premenopausal period. In- nocent tumours only infrequently arise after the menopause. OVARIAN NEOPLASMS 443

Malignant neoplasms seem to occur at a somewhat later period, although the average ages quoted are rather lower than those usually described. Putting aside cystic teratomata, it is, moreover, of the greatest interest to observe that for innocent and malignant neoplasms con- sidered together the average ages for all varieties fall within an eight-year period, forty-two to fifty years. We can, therefore, with confidence assert that the involutionary period of reproductive activity is that in which neoplasms are most generally seen, as is the case in respect of malignant diseases elsewhere. The fact that innocent neoplasms of the fall within this period seems an obvious pathological principle in view of the special life history of the organ concerned. Although in this section we have dealt almost entirely with average ages, it must not be forgotten that many cases of all kinds must inevitably fall above and below the figures given; and since the age of the patient is always taken into account, it is those that are outside the average period which often cause difficulties in diagnosis and treatment, as will become evident directly, so it is never wise to lay too much stress on average figures: if good results are to be obtained, each case must be judged on its merits. PARITY Table IV shows the number of cases of each type of innocent tumour considered in relation to parity. TABLEIV Innoce~ttNeopla.sms in relation to Parit?/

I I Type of Ncoplasrn Cystadenonra Papillary I Derrnoid Fibroma Total nunlher of cases una- lyaed * ...... 212 58 ------Percentage: Single...... 27.2 Married ...... 72.8 Nulliparous ...... 33.9 Parous...... 66.1

* Cases recorded by Smith (50), Fleming (48), and ourselves. It will immediately be noted that, in spite of text-book state- rnents that the question of parity is irrelevant, the above figures show a large preponderance of parous patients among women with TABLEV Symptonzatology of O~?ariani\+eoplasms*

Innocent Neoplasms

Ascites Dysmenorrhoea 3Zemtrual irregularities Postmenopausal and metrostaxis patients Number Abdominal Abdomi- Backache Diagnosis of cases swelling nal pain (per analysed (per cent) (per cent) cent) Number Number Sumber of cases Per cent of caacs Per cent of ca~es Per cent Per cent analysed analysed analysed

-pp-pp Cystadenomsta.. . . 174 63.2 44.8 11.5 116 5.1 116 21.5 212 30.7 35.0 Papillary cysts.. . . . 93 34.4 23.6 19.3 74 4.0 74 37.8 95 38.0 31.0 Dermoids...... 117 11.9 11.9 1.7 117 0 20 30.0 122 67.0 14.6 Fibromata...... 52 29.0 54.0 36.5 71 24.0 1 18 33.0 58 39.0 26.0

Malignant Neop!asms

Carcinomata (Lepidomata) . . . . 161 60.8 48.7 - 136 48.5 - - 161 41.0 40.3 Sarcomata ...... 15 86.6 80.0 - 15 80.0 - - 15 - 66.0

* Cases recorded by Smith (50), Fleming (48), Briggs snd Walker (46), Whitehouse (52), and ourselves. OVARIAN NEOPLASMS 445 innocent tumours-a majority so great, indeed, that it can hardly be due to a statistical error owing to the smallness of the number of cases analysed; and it certainly bears no relation to the age of the patient since the average age has already been shown to be high. One hundred and sixty-one cases of carcinoma investigated by Smith (50), Fleming (48), and ourselves show that 18.6 per cent of the patients were single, 33.3 per cent nulliparous, and 66.7 per cent parous. It is important always to exclude single patients in the estimation of parity, unless it is known that they have been preg- nant. Here again we find that a large majority of the patients falls within the parous group; and, as with innocent neoplasms, since the average age is at the end of the reproductive period of life, there can be no statistical error on account of age. SYMPTOMATOLOGY In regard to symptomatology it is again difficult to find uni- formly recorded series of cases; consequently we have summarized in percentages. the principal points noted, and have indicated the number of cases from which the percentages are estimated. We also give in Table VI 50 of our own cases of malignant lepidomata, including "endotheliomata," in which the clinical features are set out in connexion with each case, the percentage results being set forth at the foot of each column for comparison with the figures given in Table V for malignant neoplasms, which include cases recorded by others as well as by ourselves. The physical sign of ascites is inserted for comparison with "abdominal swelling." Menstrual Disturbances: So far as menstrual disturbances are concerned, too much attention should not be directed to them, for such conditions as dysmenorrhoea and menorrhagia may in many cases be associated not with the ovarian lesion, but rather with coincidental unrelated conditions of the . Ascites: The accumulation of ascitic fluid must, of course, .be related to swelling of the abdomen, of which patients may complain. Free intraperitoneal fluid in association with ovarian neoplasms is present only in certain well-defined circumstances: (1) The tumour is solid in structure and is free and mobile above the pelvic brim. It may be innocent or malignant. The ascites then appears to be caused by mechanical irritation, for similar tumours within the true pelvis, and relatively immobile, are not themselves causal factors in the production of ascites, even though it be present in malignant cases. TABLEVI: Clinical Features of Malignant Neoplasn~sof the Ovary

Bleeding or P~tmso-Unilateral Initials Diagnosis Age Para Ascites m,","g~Ual Metnstases or Abdp",pl Backache disturbances bilateral ----- J.A.S. Primary 46 5 Slight at 0 0 0 Unilateral + 0 carc~noma operation ~~~~~ L.L Carcinomasec- 62 5 ff + + On peritoneum Bilateral 0 0 ondary to car- cinoma of the treated with radium ------W.M. Primary papil- 52 0 + + 0 + On peritoneum, Bilateral 0 0 lary carcinoma omentum, ? liver ------M.K. Carcinoma 44 Si~tglu 0 0 0 0 Bilateral 0 + secondary to muvary CUrCltlOm8 ppp--p -- E.S. Primary 48 1 0 0 0 0 Bilateral + 0 carc1ooma ------B.J.P. Carcinoma 52 0 3-+ 0 - Omenturn, peri- Bilateral 0 0 secondary to toneum, surface m-ry of bowel carctnoma ------C.H. Primary pnpil- 51 0 ff 0 + Peritoneum Bilateral + 0 lary carcinoma ------M.C.W. Carcinomasec- 55 0 ++ 0 + Nodules on liver Bilateral + 0 ondary to car- cinoma of cer- vix treated with radium -- -- J McC. Primary 47 0 ++ 0 0 Perito~ieumand Bilateral 0 0 carcinoma liver H.D. Primary 44 Si~~gle f 0 0 Peritotreum sur- Bilateral 0 0 carcinoma face of bladder --- E.D. Primary 53 4 0 0 + All over abdo- Bilnteral + + carc~noma I~CII - E.L.H. Primary 48 Si~~gle S + + 0 0 U~~ilatcral 0 0 carcuioma ------A.H. Pri~tlnrypapil- 44 0 + 1- 0 0 Orner~tu~nand Bilataral 0 0 Inry carcit~omu peritu~ieum ------~ -- A.S. Pri~nary 54 Sii~~le 4- i- 4- 0 Bilateral + 0 carctnoma -- -.------E.D. Caroinomasec- 33 1 0 0 0 0 Bilateral f f ondary to rec- tal carcinoma --- J.E. Primary 59 0 0 + + 0 Unilateral + 0 oarc~noma L.J. Carcinoma 35 0 0 0 0 0 Bilateral 0 0 secondary to stomach ---- A.E. Primary 70 1, miscar- 0 0 + 0 Unilateral 0 0 carc~noma rlage P------M.J.R. Primary papil- 46 1 0 0 0 0 Bilateral + + lary carcinomn -- - - M.C. Primary papil- 60 7 + + 0 + On bowel Bilateral + 0 lary carcit~oma

pppp-- M.D. Pri~ry 58 0 ++ 0 + On peritoneum Bilateral + 0 wclnoma ------S.W. Primary 00 0 0 0 + 0 Bilateral f + carclrloma TABLEVI: Clinical Features of Malignant Neoplasms of the Ovary (continued)

Bleeding or other Poetmeno- Unilateral IniW Dignosis Age Para Ancites Metsstasaa or Abd~y*Backaohe menstrusl bilateral -----disturbanaes --- E.M. Primsly wci- 42 1 ++ + 0 Peritoneum. Bilateral f 0 noma in cyst- liver, omentum adenoma ------L.B. Carcinomaaeo- 58 0 0 0 + 0 Unilateral ondary to car- clnoma corpus uteri --- M.F. 40 1 +f 0 0 Peritoneum Bilateral + 0 zizma --- M.8. Primary carci- 35 0 ++ 0 0 0 Bilateral 0 0 noma in cynt- adenoma -- M.B.A. Carcinomaem- 50 2 ++ 0 + 0 Unilateral + 0 ondary to car- cinoma of pel- vic colon ------E.F. Carcinoma aeo- 58 3 ++ Occssional + Peritoneum Bilateral + + ondary to car- haemor- cinoma in pel- rhea vic colon -PPP------H.L. hrywci- 65 2 0 0 f 0 Unilateral + 0 noma in cyst adenoma E.T. Prin?ary 25 1 ++ 0 0 0 Unilateral 0 0 wcmoma --- M.E.M. Primsly carci- 68 6 + 0 + 0 Bilateral + 0 noma m cynt- edenoma --- P.M. Carcinomaem- 55 12 + 0 f 0 Bilateral + 0 ondary to gas- tric carcinoma -pppp------H.S. Pry 59 0 0 + + 0 Bilateral 0 0 carcinoma ----- E.O. Caroinomaseo- 45 6 Sf 0 0 0 Bilateral f 0 ondary to car- cinoma of rea- tum ------M.B. Primary carci- 52 0 ++ 0 + In liver Bilateral + 0 noma in cyst adenoma pppp-- --- M.E. Malignant 44 0 ++ 0 0 On peritoneum, Bilateral f 0 folliculoma ----- ? in liver E.A. Hrukenbrg 32 6 0 0 0 0 Bilateral 0 0 tumourmth pregnancy eaaondaryto wtric carci- noma ----- J.P. Carcinoma 34 Single + 0 0 Bilateral 0 0 secondary to brwt M.W. Beoqndary 56 7 ++ 0 + Omentum Bilateral + + caramnoma H.L. hw,pdsry 50 3 ++ 0 Peritoneum, Bilateral caromnoma,

B.P. Privy 35 1 0 0 Omenturn-and Bilateral carcinoma ----- peritoneum E.M.P. VY 45 0 ++ 0 Peritoneum Bilateral wclnoma 448 W. BLAIR BELL AND M. M. DATNOW

TABLEVI: Clinical Features of Malignant Neoplasms of the Ovary (continued)

Bleeding or Unilateral Initials Diagnosis Age Para Ascites m,",,"&,l P","i:z"- Metaatasea or Backaclle disturbances bilateral --- M.G. Primary 65 0 ff 0 + Peritoneum Bilateral 0 0 carcinoma --- C.N. Krukenberg 42 7 ff 0 0 Liver Bilateral + 0 tumow sec- ondary to gas- tric carcinoma H.B Primary earci- 47 0 + f 0 0 0 Bilateral + 0 noma in cyst adenoma ------M.M. Primary 51 0 0 0 + 0 U~iilateral 0 0 carcinoma ------E.J. Primary 45 11 0 0 I'eritoneum Bilateral + 0 carcinoma Unnamed Primary carci- 50 9 0 0 f 0 Unilateral museum noma ln der- specmen moid cyst - -.------W. Round cell 17 0 0 0 0 0 Unilateral anrcoma ------S B.H. Malignant 0 Bleeding f 0 Unilateral + 0 folliculoma 521 Pnrous Bleeding Total number of Average excluding and Poetmeno- Bilateral A!$zlir'nl Backache caw analyeed age 5 single menstrual pausal patients disturbances ~~~~~ --- 50 49 60 6'2 14 52 76 56 14 per cent per cent per cent per cent per cent iwr cent per cent

(2) Innocent papilliferous cysts from which growth has escaped, and malignant tumours from which the neoplasm is fungating-both of which conditions lead to implantations on the peritoneum-give rise to ascites owing to the production of some irritating secretion by the neoplastic cells. An appreciation of these circumstances enables one to explain the absence-unexpected by some clinicians-of ascitic fluid in those cases in which the malignant disease is confined to ovaries fixed in the true pelvis, and in which there is no fungation or peri- toneal implantation, and so no mechanical or chemical irritation.

DIAGNOSIS When the history of the case has been recorded, special atten- tion is next directed to the age of the patient and her previous illnesses, if any; for it is most important at the outset that the biological nature of the neoplasm should be determined, so far as this is possible, by clinical investigation. This aspect of diagnosis alone will be considered here. OVARIAN NEOPLASMS 449

In a woman about or after the middle period of life the previous loss of one breast, even though exact pathological evidence con- cerning the lesion then present be not forthcoming, followed by the appearance, it may be some years later, of an ovarian neoplasm is strongly suggestive of metastatic developments. A case of this kind was mentioned in the earlier part of this communication (p. 47). We have seen so many similar cases among the patients sent to us for treatment with lead some time after excision of the breast had been practised, who had no symptom referable to the pelvic tumours, that it may be asserted that, unless routine exam- ination be made, secondary ovarian growths may escape recogni- tion for a long time. Although the patients affected with malignant disease of the ovaries are usually over forty-five years of age, it is not at all un- common to meet with the disease in women "in the thirties"; con- sequently we must always be prepared to disregard the age of the patient, or at any rate look upon it as merely of relative importance, and even that chiefly in respect of women over forty-five years of age. The physical signs, however, are of great help to us in arriving at a diagnosis. Bilateral solid, or partly solid, tumours, nodular in character and firmly fixed, especially when associated with as- cites, pain, and loss of weight, even without the supporting evi- dence of age or disease elsewhere, are practically always malignant neoplasms. Even so, we occasionally meet with cases that are exceptions. We have already referred to the curious microcystic character of some pseudomucinous cysts (p. 32 and Fig. 30), and we may use this type as one of the rare examples of an innocent neoplasm with certain malignant clinical features which throw considerable doubt on its nature until an histological examination has been made. M. S., aet. thirty-five years, was single and nulliparous. Her menses were regular, 4/28. A tumour of the left ovary had been removed by Dr. Leith Murray two and a half years previously. We are informed by him that in ap- pearance the tumour resembled exactly that now to be described, and that there was free fluid in the peritoneal cavity. The general condition of the patient was good at tJhe time, and the right ovary had been in- spected and considered normal. There had been some doubt as to the nature of the tumour, but this had been resolved after an examination of sections, when a diagnosis of pseudomucinous cystadenoma had been made. 450 W. BLAIR BELL AND M. M. DATNOW

Two and one-half years later the patient was admitted to the Liver- pool Royal Infirmary under the care of a , Professor John Hay. She then complained of swelling of the abdomen for twelve months, to- gether with loss of weight and swelling of the ankles. Fluid was detected in the right pleural cavity, from which 11 pints of clear fluid were with- drawn. A diagnosis of free fluid in the peritoneal cavity was also made, and 8 pints were evacuated. A fixed tumour was then felt in the abdom- inal cavity, and the case was referred to us for treatment. The diagnosis of malignant ovarian tumour had been made and was supported by our examination. At that time we had no knowledge of the nature of the neoplasm previously removed, although we received information that the left ovary had been excised for a tumour. The operation was performed under spinal anaesthesia. The right ovarian tumour shown in Plate V, which was adherent, was removed, together with a considerable quantity of free ascitic fluid and also pseudo- rnucinous secretion which seemed to be oozing from the tumour in several places, as does honey from a honey-comb. No evidence of any other tumour was found. The patient rn:tdc tin excellent recovery and has remained well for two years. On histological examination a diagnosis of microcystic pseudo- mucinous cystjadenoma was made (Fig. 30).

Here then was an example of a tumour with most of the signs and symptoms of a malignant ovarian tumour. How much pain the patient suffered was difficult to determine in view of her general discomfort . Innocent papillary tumours are not often associated with loss of weight, nor indeed with pain, but ascites may be considerable owing to peritoneal implantations. On bimanual examination, if the papillary growths have escaped from the cyst in which they originated, the mass formed is softer than that of a malignant tumour. Difficulty may arise in all those cases in which we are confronted at operation by equivocal appearances such as we have already recorded (p. 9) and to which we shall again refer directly. In this connexion, mention may be made of sarcomata of the ovaries. It is in relation to the differences in appearance between these and fibromata that we are immediately concerned, and were indeed questioned in the discussion at the British Congress of Obstetrics and Gynaecology (44). We shall, therefore, repeat the remarks made by one of us in reply. Sarcomata, including en- dotheliomata, are not so uniform and pale in colour as fibromyo- mata, for often there are areas of haemorrhage, and the sarcoma- tous tumour is, therefore, usually much the softer of the two. M. S., AET. THIRTY-FIVE \-EARS: BISECTEDMICROCYSTIC PSECDOMUCINOUSCYSTADENOMA, ASSOCIATED WITH CLINICAL FEATCRESSUGGESTISG A RIALIGSANT LESION. X $i

OVARIAN NEOPLASMS 451

This, however, does not neccssttrily apply to the hard, spindle-cell variety. Again, sarcomatous tumours are usually adherent at some part of their surface, if not extensively. Fibromata are nearly always quite free, or adherent only to a very limited extent, unless the tumour has undergone degeneration, and this is not uncommon in the larger fibromatous tumours. Pain is absent from undegenerat'ed fibromata, but present with sarcomata. Uncomplicated innocent cystic tumours in young women are diagnosed without difficulty, even when bilateral. Those cases, however, which in elderly women become malignant, may at first have but few signs and symptoms significant of the change that has occurred. Nevertheless, not infrequently, we are led to suspect the sequence of events by the knowledge that the patient has known of a large tumour which has existed for many years, but has given little trouble or cause for anxiety. Latterly, she may tell us, the tumour has increased in size, and there has been constant pain referable to it. The patient may have begun to go "down- hill" also. We attach great importance to pain in these cases. I11 TREATMENT The treatment to be applied to ovarian neoplasms of all kinds is chiefly surgical in character; but the actual procedures practised depend for the most part on the biological nature of the growth. In those cases, therefore, in which its character is not manifest before operation, a consideration of the naked-eye appearances of the parts as disclosed after laparotomy becomes imperative; yet, as we have already recorded (p. 8), even then it may not be possible for the operator to be quite certain. When, however, there is knowledge of a primary growth, coexistent or previously treated, in the breast or gastro-intestinal tract, little doubt can arise as to the nature of the ovarian lesion, however misleading certain fea- tures may be, as in the following case: Mrs. A. B., aet. forty-one years, was a primipara. Her menses were regular, 5/28. The right breast had been removed two years previously for carcinoma mammae. The patient was in good condition, and com- plained only of sacralgix and pain down the thighs. Examination revealed :I nlovable cystic t,umour on the right side of the true pelvis. In view of the previous history, this was diagnosed as being probably secondarily malignant in spite of its resemblance on bimanual examin:ttion to an innocent cystoma. At operation the specimen shown in Plate VI was removed. In accordaice with our usual practice in malignant disease of the ovary, 45 452 W. BLAIR BELL -4ND M. M. DATNOW the uterus and the appendages on both sides were excised. Histological examination of a section through a nodule in the cyst wall of the right ovary (Fig. 58) showed the nodule to consist of a mass of malignant spheroidal cells, while the cyst wall was lined by the cells of the Graafian follicle:-the tumour was a large follicular cyst in the walls of which metastases had occurred. The left ovary cont)ained no metastasis, but the nodule on the left infundibulopelvic ligament, noted at operation and removed, was malignant. The patient made a good recovery and remained well without any

recurrence for more than one year, when she succumbed to an acute in- fection. Here then were implantations which might have escaped notice for some time had there not been an innocent cystic condition of the right ovary in which the secondary growths were incor- porated. In contrast with this case the following is of interest: Mrs. C. W., aet. 54 years, had had five children and seven miscarriages. The menopause had supervened seven years earlier. The previous history of the patient contained nothing suspicious of malignant disease. The present symptoms consisted of bearing-down pain which had been of several years' duration, but had been getting worse for some months. The general condition of the patient was very poor. PLATE VI

A. R., AET. FORTY-ONEYEARS: XODULES OF SPHEROIDAL-CELLCARCINOMA ARE TO BE SEENIN THE WALLSOF A FOLLICCLARCYST OF THE RIGHTOVARY. X There is a small malignant nodule on the left infundibulopelvic ligament. Com- pare with Plate VII.

PLATE VII

C. W., AET. FIFTY-FOURYEARS: SOLID NODULES OF FIBROUSTISSUE ONLY ARE TO BE SEENON THE WALLSOF FOLLICULARCYSTS OF THI OVARIES. X 3/2 Compare lvith Plate VI.

PLATE VIII

11. C., AET. THIRTY-TIIREEYEARS: PARTLY DEGENERATED FIHROMA OF THE OVARY, GROWINQ AWAYFROM -4 COSSIDERAHLE;\MOUNT OF NORMALOVARIAN TISSUE WHICHCOULD HAVE BEEN CONSERVED BY ENUCLEATIONOF THE TUMOUR.X 1

OVARIAN NEOPLASMS 453

On examination, bilat,eral cystic tumours, in which hard nodules could be felt, were discovered in the true pelvis. Malignant ovarian tumours were diagnosed in view of the age and condition of the patient. No primary tumour elsewhere could be located. 1,aparotomy was per- formed, and the specimen shown in Plate VII was removed. This has all the appearance of cystic ovaries with malignant nodules; consequently the appearance of the tumours, the age of the patient, and her poor con- dition left no doubt in our minds as to the correct procedure to adopt-a " clean sweep."

FIG. 59. C. W.: SECTIONTEIROUQH FIQ. 60. C. W.: SECTIONTHROUQH THE WALLOF A CYSTIN A CYSTICOVARY A NODULEON THE WALL OF A CYSTIC WITH SOLIDNODULES (PLATE VII), HHOW- OVARY(PLATE VII), SHOWINGFIBROUS INQ A LININQ OF THE MEMBRANAGRANU- TISSUEONLY (SAMECASE AS Fro. 59). LOSA. X 222. X 81.

In no section of many made was any evidence of malignancy forth- coming. The cysts were apparently follicular in type, in spite of the age of the patient, being lined with cubical cells (Fig. 59). The hard nodules were composed of fibrous tissue only (Fig. 60). The patient has remained well for three years.

Once more, therefore, we see how extraordinarily deceptive macroscopic evidence can be, whatever may be the coincidental conditions of age and general health.

We shall, however, now consider the treatment of tumours be- lieved to be innocent and afterwards discuss the management of those considered to be malignant, recognizing always the fact that careful histological examination alone subsequently to removal can decide their nature in certain cases. 454 W. BLAIR BELL AND M. M. DATNOW

Innocent Neoplasms The treatment of these is surgical only. We shall not attempt to deal with the surgical technique in general, but shall mention one or two principles which seem to be important, yet often overlooked. First, as to the route: it does not appear to be generally ap- preciated that it is of great advantage to the patient if the tumour or tumours be removed through the posterior vaginal fornix. When the tumour is free in the true pelvis, and especially when the patient is young, multiparous, and pregnant, this should be the route of election. To those experienced in vaginal technique any procedure deemed advisable, such, for example, as partial resection of the ovary, is a simple matter. If the tumour be cystic or partly cystic in character, it is per- fectly safe and easy for the surgeon to fix it in the pouch of Douglas and evacuate the contents into the before delivering the collapsed cyst wall. In the presence of extensive adhesions, free papillary growths in the pelvis, and malignant lesions, the vaginal route is contrain- dicated. When the abdominal rohte is chosen, as it should be for turn- ours above the brim of the true pelvis and for those in which the vaginal route is contraindicated, the question arises concerning the dimensions of the laparotomy incision to be made: Bland- Sutton was probably the first to advise the removal of all cystic ovarian tumours entire, rather than by first tapping and then re- moving the sac. He stated that most bilateral tumours were malignant or papillary, and that, therefore, there was always the potential danger of implantation should tapping be practised. His premises, however, are now known to be less gener:~lly true than he supposed. Formerly, very large cysts were removed through the small two-inch incisions, beyond the limits of which the earlier abdominal surgeons feared to go, for in those days the small abdominal incision was regarded as being a special factor in safety. The pendulum, however, may have swung too far in the op- posite direction. illany hold that it has, and that there is no necessity to open the abdomen from sternum to pubis to remove a very large cyst. We ourselves have always favoured the large incision, partly t~ecausethe larger the cystadenoma in persons past middle age the more likelihood there is of it being malignant in some part, and the OVARIAN NEOPLASMS 455

greater the frequency of omental and intestinal adhesions, which are more easily separated if the cyst be untapped. No doubt there are exceptions, for on the one hand the most common cystic neoplasm is the innocent cystadenoma, and the majority of these are not malignant, nor are the papillary outgrowths they may contain usually transplantable. But, on the other hand, we have often been thankful we have removed these tumours in- tact for subsequently we have found evidence of malignant neo- plastic changcs in them. It is unfortunately true that but few surgeons give any con- sideration to the surgical treatment of innocent neoplasms other than the complete removal of the affected organ. Yet we are convinced that this attitude is unjustifiable, and is not in accord with the trend of scientific . One of us (43) has recently discussed the present position of conservative surgery generally, and has shown how much can be accomplished provided that all premenopausal lesions, which are not malignant, be approached in a physiologically reverent and conservative spirit: eradication should never be practised without the surgeon feeling the humilia- tion of defeat. The complete removal of both ovaries for neo- plasms during the reproductive period is rarely necessary, as we have already indicated (see Case 2, p. 9). As a rule, some healthy portion of ovary can be preserved with its normal connexions, for nearly every neoplasm tends to grow away from the hilum (Plate VIII), and the preservation of this with adjacent ovarian tissue leaves well nourished and functional tissue. Moreover, it has been shown by one of us (42) that follicular secretion is not necessary to maintain the integrity of the uterus and that the interstitial tissue is sufficient to do so. When, however, both ovaries are removed and it is impossible for the surgeon to save ovarian tissue in'situ, either because the blood supply has been affected, or there is some doubt as to the exact biological nature of the tumour, an ovarian graft in the rectus muscle should be made to maintain genital activity in a young patient. Should malignant disease be ultimately found on his- tological examination, no harm will have been done, for the trans- planted tissue cttn immediately be excised. However, in the circumstances mentioned it is not often that real doubt exists, and it is easy for the operator to remove the tumour and preserve normal ovarian tissue. We would call attention here to a fallacy which is often ex- 456 W. BLAIR BELL AND M. M. DATNOW pressed: namely, that when one ovary only is affected at the time of operation it is quite unnecessary for the surgeon to take the trouble to conserve any portion of the affected organ, for a normal ovary is to be left, and this is enough. It may be for the time being; but how often the second ovary is removed later for a lesion similar to that for which the first was excised. In innocent bilateral papillary cysts with peritoneal implanta- tions it is sometimes but not always necessary to remove both ovaries. The disappearance of the peritoneal implantations may follow removal of the parent growth. We can submit no evidence on the point, but we see no reason to doubt that the same result would follow conservation of a healthy portion of an ovary, or :in ovarian transplantation. About and after the menopause it is always wise, even with unilateral, innocent, or apparently innocent, tumours, to remove both ovaries and the uterus supravaginally in nulliparous, and completely in parous women. In healthy patients there should be no added operative risk. As we have already said, malignant disease not uncommonly supervenes during the later periods of life in innocent cystic tumours which have been know to exist for many years. Malignant Disease of the Ovaries The treatment of malignant disease of the ovaries is fraught with difficulty in a majority of all cases. It may be definitely asserted in respect of surgical treatment alone that a good prognosis can be given only when the disease, whether primary or secondary, is limited to the ovary or ovaries, and no adhesion is present. In these circumstances a cure is often effected by removal of both ovaries, Fallopian tubes, and the uterus. The greatest care must, however, be taken lest the affected organ or organs be damaged and ('cellular spill" occur. We must, too, call attention to the fact that, although the malignant disease appears to be limited to one ovary, it is dis- astrous to remove that organ alone and to leave the uterus and the other ovary. The following case illustrates the rapidity with which the second ovary may become involved: H. D., aet. forty-four years, menstruated regularly 24/26. On March 28th the left ovary had been removed elsewhere for a tumour which was found to be malignant-primary adenocarcinoma. Some small nodules in the pelvic peritoneum were noted. The right ovary appeared to be OVARIAN NEOPLASMS 457

normal. When seen on May 3rd in the same year-only six weeks later -on the right side of the pelvis "a rounded tumour of the same size as that of an orange" with nodules in the pouch of Douglas was discovered on examination. The ovary, Fallopian tubes and uterus were removed, but the patient died of metastases twelve months later. When an adhesion is present the prognosis immediately becomes grave, for malignant cells can often be demonstrated passing through the newly formed tissues. Many or extensive adhesions add to the danger. The following case affords an unfortunate example of what may happen even with a single small adhesion. Miss C. B., aet. tjhirt,y-nine, w:ts operated on in October 1916 for the rcrnoval of bilateral primary adenocarcinoma (Fig. 61) of the ovary.

FIG. 61. C. B.: SECTIONSHOWING FI~.62. C. B.: SECTIONSHOWING J'RIMARY ADENOCARC~NOMAOF THE OVARY PLUGOF CARCINOMACELLS IN AN AD- ENCLOSEDBY NORMALOVARIAN TISSUE. HESION (SAMECASE AS FIG. 61). X 225. X 22.

Both tumours were completely free except at one point, where there was a limited adhesion between the right ovarian tumour and the peritoneum at the pelvic brim. This piece of peritoneum, which was about 1 inch in diameter and overlay the right ureter, was carefully excised with the tumour. The separated edges of peritoneum were sutured together. In a section of the adherent portion of peritoneum penetrating cancer cells were seen (Fig. 62) The patient remained well for nearly two years, when involvement of the mediastinal glands appeared, and she died some months later.

When the disease is not limited to the ovary, whether the spread be direct from a primary or secondary growth, or part of a general 458 W. BLAIR BELL AND M. M. DATNOW peritoneal dissemination, surgery can be but palliative. This, however, should not deter the experienced surgeon, whose enthu- siasm will not carry him too far. We have seen great relief follow the removal of both ovaries and the fundus of the uterus not only in the allevit~tionof pressure symptoms, but also in the diminution or even permanent disap- pearance of ascitic fluid. Moreover, the excision of such malig- nant masses facilitates the subsequent treatment of the remaining disease by chemotherapeutic measures. In a large proportion of the patients who suffer with mn1ign:int disease of the ovaries the general condition is poor. The patient is often elderly and unfit for prolonged operation or for inhalation anaesthesi:t. We have found that preliminary blood-transfusion, spinal anaesthesia, and injections of infundibular extract to main- tain the blood pressure at the normal level, greatly lessen the danger of operation. . Apart from surgery or in association therewith, radiation and chemotherapy, especially with lead, have been tried. 1Gnough has been said to show that in all operable cases surgical tre:ttment has a foremost place. Unhappily, however, surgery like has but local application, and in the cases under tliscussion the localization refers to the ovaries and uterus alone and not even to the neighbouring peritoneal tissues. Radium has no place in the treatment of this disease, and in our experience x-ray treatment does more harm than good. With lead we, Reckwith Whitehouse (52) and others have had much encouragement, especially when the advanccd nature of the disease in the cases treated is taken into consider at' ion. From what has been said it will be evident that considerable experience, with the judgment that follows, is essential when the surgeon deals with malignant disease of the ovaries, for the trest- ment is rendered particularly difficult by the anatomical re1a t'ions of these organs, by the manner of spread, and by the frequency of coincidental lesions elsewhere.

SUMMARY 1. It has been shown that the previous history of the patient is of considerable importance, especially in respect of malignant neoplasms. 2. The average age of patients with both innocent and malig- nant neoplasms, except cystic teratomata, and possibly ~arcornat~:~, OVARIAN NEOPLASMS 459 falls between forty-two and fifty years-the period of reproductive involution. 3. A large majority of all women with ovarian neoplasms, in- nocent and malignant, is parous. 4. The symptomatology is somewhat ambiguous in that men- strual disturbances may be due to associated uterine lesions rather than to the ovarian neoplasms. Abdominal pain and loss of weight, however, are important indications of malignancy. The physical sign of ascites cannot be relied on as a determining facttor in the diagnosis since it is frequently present with innocent solid, and occasionally cystic, ovarian tumours. Ascites is observed only in certain well defined circumstwnccs :ind is due to mechanic:~l or chemical irritzttion of the peritoneum. ti. The diagnosis is considered here only from the point of view of the biological nature of the neoplasm. 6. Treatment is entirely surgical, and should always be prac- tised in cases of innocent neoplasms. With these lesions conserva- tion of normal ovarian substance should be attempted in young women. Difficulties in diagnosis may sometimes deter the surgeon from acting in this manner. In parous women about or after the menopause, who form thc majority of all cases, bilateral salpingoophorectomy with panhys- terectomy should be practised. If the patient be nulliparous, supravaginal with salpingoophorectomy is sufficient. In young parous wornen the removal of ovarian cysts by the vaginal route may be practised with advantage. If laparotomy be performed, the tumour should be removed without preliminary tapping or puncture. The treatment of malignant neoylasrns is curative or palliative. Lead therapy is often a useful adjunct to surgical procedures. Radiotherapy is not to be recommended.

42. BELL,W. BLAIR:Sex Complex, Baillikre, Tindall, & Cox, London, 1920, 2nd ed. 43. BELL,W. BLAIR:Brit. M. J. 1: 653, 1931. 44. BELL,W. RI,AIR:J. Obst. & Gynaec., Brit. Emp. 38: 386, 1931. 45. BRIDE,J. W.: Trans. Edin. Obst. Soc. 1929-30, p. 32. 46. BRIGGS,H., AND WALKER,T. E.: J. Obst. & Gynaec., Brit. Emp. 13: 77, 1908. 47. ERDMANN,J. F., AND SPAULDING,H. V.: Surg. Gynec. & Obst. 33: 362, 1921. 460 W. BLAIR BELL AND M. M. DATNOW

48. FLEMING,AMY M.: J. Obst. & Gynaec., Brit. Emp. 35: 280, 1931. 49. MCINTYRE,D.: J. Obst. & Gynaec., Brit. Emp. 38: 302, 1931. 50. SMITH,G. VAN S.: Am J. Obst. & Gynec. 18: 666, 1929. 51. STEVENS,T. G. : J. Obst. & Gynaec., Brit. Emp. 38: 256, 1931. 52. WHITEHOUSE,B.: J. Obst. & Gynaec., Brit. Ernp. 38: 264, 1931. 53. WILLIAMSON,H., AND BARRIS,J. D.: New System of Gynnecology, by Eden and Lockyer, 1917, vol. 2, p. 769.