G. L. Foss
O.B.E., V.R.D., M.A., M.D. Camb.
CLINICAL ASSISTANT TO THE ENDOCRINE CLINICS, UNITED BRISTOL HOSPITALS.
IT has long been known that the ovaries influence the growth of the breasts, and that oophorectomy sometimes causes breast cancer to regress (Schinzinger 1889, Beatson 1896, Boyd 1900, Morris 1902, Thompson 1902, Lett 1905a and b).
Haddow (1935) found that certain carcinogenic hydro-carbons retarded experimental malignant growths in laboratory animals, and some of these substances had an oestrogenic effect. This discovery led to a trial of synthetic œstrogens for carcinoma of the breast. Most workers agreed on the value of œstrogens in postmenopausal women and that the best responses were obtained in soft-tissue métastasés, with little benefit in skeletal métastasés.
Androgens were given to patients with breast cancer to inhibit œstrogens which might be stimulating the growth (Loeser 1938, 1939, 1940, MeWhirter et al. 1944, Adair and Herrmann 1946, Herrmann and Adair 1946, Cutler and Schlemenson 1948, Prudente 1948, Lowen- haupt and Steinbach 1949, Douglas 1952, Hallberg et al. 1953).
It is now considered that some breast cancers are dependent on œstrogens, whose removal will improve the prognosis. This view is emphasised' by Cade (1955) and Dodds (1955).
The insufficiency of oophorectomy and/or androgen therapy gradually became evident. Œstrogens secreted by the adrenal cortex are increased after oophorectomy because tbe adrenal glands then hypertrophy. With the advent of cortisone to control patients during and after operation it was logical to remove the adrenal glands. Since Huggins and Dao (1952) and Huggins and Bergen- stal (1952) reported a series of cases so treated, adrenalectomy and oophorectomy have been widely practised for the young patient with advanced métastasés. If extirpation is complete, only those cases which presumably depend on œstrogens show a rapid and dramatic improvement and a recession of tumour growth which may last several months.
All workers are agreed that there is no means of determining which case depends on œstrogens, and can then he expected to improve under treatment/ except trial of hormone therapy. This takes several weeks, hut there is a dramatic response one to six days after adrenalectomy if the tumour is oestrogen-dependent.
Old postmenopausal patients often show a spectacular temporary remission, particularly of soft-tissue lesions, after large doses of œstrogens—-possibly owing to a local effect as well as to inhibition of the anterior lobe of the pituitary gland.
Materials
This paper is an analysis of 106 cases of breast cancer treated with either oestrogen or androgen in women who died in a period of seven years, and the following data relate to the patients’ state when accepted for hormone therapy.
All the patients had advanced stage-IV cancer of the breast and had already received all possible benefit from either X-irradiation or surgery or both, or else the lesion was too far advanced for anything except palliative hormone therapy.
Previous treatment.—Of the 106 women 3 had had surgery alone, 10 X-irradiation alone, 82 surgery and irradiation, and 11 no treatment.
Histology.—78 patients had histologically proved carcinoma, but Broders’s grading was not stated in many of the pathological reports. Kecords of histology were missing in 7. There were no pathological reports in 21 cases which were much too advanced for surgery and in which the diagnosis was obvious on clinical findings.
The duration of disease before the start of hormone treatment was as follows : 94-5% of the women had had the cancer for 6 months, 86-0% for 1 year, 68-0% for 2 years, 48-0% for 3 years, 34-0% for 4 years, 21-0% for 5 years, 10-4% for 7 years, and 4-0% for 10-15 years.
Connubiality.—24-6% of the patients were single and 75-4% married.
Age-distribution.—9 patients (8-5%) were aged less than 40, 61 (57-5%) were aged 41-60, and 36 (34%) were aged 61 or more.
The menopause had occurred in 81%, and of these it had followed either irradiation or oophorectomy in 12-6%.
Pregnancy or miscarriage was recorded in 55-6%.Some patients had métastasés in several or all of these sites.
Distribution of métastasés was as follows :
Table 1.
General condition.—Advanced cardiovascular disease was found in 8-5%. The general nutrition was good in 37-8%, but 56-5% had lost weight up to 2 st. and 5-7% more than that.
Methods
All the patients were interviewed and examined monthly or every two months according to their condition and response to treatment. Serial records of weight, haemoglobin, erythrocyte-sedimentation rate (E.S.R.) (Westergren), and blood alkaline and acid phosphatase were kept. Changes in lesions were recorded photographically or followed on serial radiographs.
Estrogen Therapy
Originally the British Empire Cancer Campaign wished to compare the effect of low dosage (0-3 mg. q.i.d.) with that of high dosage (5 mg. q.i.d.) of stilbcestrol. Printed forms were received, and patients were allocated at random to the dosage printed on these ; by this means no individual selection of high or low dosage was made by each investigator. Patients were, if possible, kept on their allotted dosage until death- A change to high dosage was made only if it was obvious that no effect was being obtained from low dosage. Similarly, on occasions the dosage had to be lowered if symptoms of toxicity were intolerable.
Much individual variation was noted in tolerance within the dosage range 0-3-5-0 mg. q.i.d. After preliminary nausea, anorexia, flatulence, or giddiness most patients became gradually tolerant, but a few remained unable to take stübœstrol. In my experience small doses of this hormone are practically valueless ; hence early doses of 5 mg. q.i.d. are advocated, with increase of this amount to as much as 50 mg. q.i.d. if required.
Although the oestrogen was usually stilbostrol occasionally ethinyl œstradiol was substituted because of intolerance or sometimes when large dosage was required.
Androgen Therapy
57 patients were treated with androgens—of these 6 were aged less than 40, 38 were aged 41-60, and 13 were aged 61 or more ; 35 of these (61-4%) bad bone secondaries when first seen.
To determine whether androgen therapy was of benefit or not, it was decided to give the largest possible tolerated dosage and to maintain this until there was some clinical or radiographic improvement, or until it was obvious that the patient was rapidly deteriorating. When this trial started, it was necessary to collect as many cases 1 possible for inclusion in the British Empire Cancer I Campaign trial of low and high dosages of oestrogen, and some of the early cases were so treated although bone secondaries were present, when better judgment would have indicated treatment with androgens. However, after some experience, the younger patients and those with osseous métastasés were given only androgens.
The high dosage chosen, 300 mg. daily, was not limited at first by expense, because adequate material was freely available through the generosity of several firms.
Most patients treated with androgens were given daily injections, but some, too thin to tolerate repeated treatment by this route, or to give them a brief rest, were given oral therapy with either methyl androsteuediol or methyltestosterone. A few were also treated with implants of testosterone or of testosterone prop1 mate for various reasons—e.g., difficulty of regular attendance and supervision.
Large dosages of androgens were usually continued indefinitely, even when patients were in a quiescent stage in reasonable health, though often the freuuency of dosage was lessened, only to be restored with signs of I further deterioration or spread.
For example, in a patient who survived for 260 weeks, I the following enormous doses were given : 60·7 g, of testo-1 sterone propionate, 84·0 g. of testosterone isobutyraie, 10·5 g. I of methyl androstenediol by injection and 8·4 g. by mouth, I and 6·72 g. of methyltestosterone by mouth up to about 41 month before death.
In the patients who had been menstruating until the I start of treatment amenorrhcea persisted.
Toxic Effects
Estrogens
The well-recognised toxic effects of stilbcestrol and ethinyl œstradiol are common even with minimal dof for endocrine dyscrasias, but (as already stated) a much I bigger dosage was believed to be necessary and given. Hence it is not surprising that 40-3% of the patients so treated complained of toxic effects. It was necessary to encourage these patients to persevere with the prescribed dosage, and usually tolerance improved. Some patients assimilated 160-200 mg. of stübœstrol daily.
Toxic symptoms noted were anorexia, flatulence, nausea, retching, sometimes vomiting, dizziness, i general malaise. These symptoms developed either early or late in treatment ; in the latter case they were possibly due to hepatic métastasés impairing detoxication. Often these patients were lethargic, dehydrated, and wasted with dry inelastic sallow skin.
Androgens
Only 29-8% of the patients treated with androgen complained of nausea or vomiting. Sometimes it was necessary to reduce the daily dosage, and potassim citrate and a powder of magnesium trisilicate and bismuth carbonate were of value. In view of the suggestion that the toxic effect in the earlier cases in this series tres with massive therapy was due to hypercalcsemia, the serum-calcium level was investigated on many occasions but was never found significantly raised.
Side-effects
Estrogens
Postmenopausal bleeding was not uncommon in patients taking a small ineffective dosage of 0-3-1-0 mg. q.i.d. for a monta ox so ; treatment had then been stopped by either the patient or the doctor. The correct procedure is to increase the dosage to 5 mg. q.i.d. and to impress on the patient the importance of not interrupting regular therapy. With 5 mg. q.i.d. interval bleeding takes place only rarely, and then perhaps after the tablets have been omitted for a day ; if patients were given continuous therapy with 20 mg. q.i.d. or more for many months, there was no bleeding. Withdrawal of oestrogen caused (bleeding in 31-2% of 61 menopausal cases : 10 of them (bled oii a dosage of stilboestrol from 0-3 mg. q.i.d. to (5 mg. q.i.d., and 9 on one of 5-0 mg. q.i.d. ; of the 42 remaining patients 2 on 0-3 mg. q.i.d. had not bled in I seven and eleven weeks before death, and no report was bio for 1 patient who discontinued attendance. In 130 patients who received 5-0-50 mg. q.i.d. there was no bleeding during their treated lifetime—-sometimes for as Ilong as 125 weeks. There was no bleeding' in 3 patients while having ethinyl oestradiol 0-05 mg. q.i.d. or more, in patient while having œstradiol benzoate 10 mg. on alternate days, and in 5 patients while having stilbœstrol 1-5 mg. q.i.d. in addition to a daily dose of 100-300 mg. of testosterone propionate by injection.
Pigmentation was striking in some patients, particularly the older group and in those who were normally darkly pigmented, and in this series it developed in 32- 8%. It 5 very intense around the nipples and areolæ, the linea nigra, and the perineum, and in senile keratoses. The colour varied from dark brown to almost black. The nipple of an irradiated breast was often less pigmented than that of the other breast. After months of therapy the pigraentation often disappeared, even though big doses were continued. In association with this there was sometimes sufficient kératinisation of the nipples to give them a greyish black tinge which, like the pigmentation, ultimately disappeared, leaving a paler smooth nipple.
(Edema of the ankles and lower legs was not unusual with continued large dosage, and sometimes the cedema- jtous arm on the side of the radical mastectomy was still further enlarged. Earely in some of these old patients congestive failure was precipitated, but it was usually only necessary to restrict the intake of salt and to give an oral diuretic—e.g., ‘ Mercloran ’ and acetazolamide.
Androgens
Side-effects with androgens naturally depend on the duration of treatment and the dosage, but there was also individual susceptibility. The most powerful known action of androgens is anabolic, with retention of nitrogen aud sodium, increasing the weight of many patients, and taking their muscles noticeably larger and firmer though the patients had advanced cancer. Much of the benefit of androgen therapy is probably due to this effect, with objective improvement of appetite and sometimes an •Acreage in well-being and energy, enabling patients to continue their work or their domestic duties. There was often pronounced plethora.
The unpleasant side-effects are those of virilisation This varied from one patient to another : some were tapidly affected, whereas others, who survived only a hort time, showed very little virilisation. The distribu- ,10A of signs of virilisation, expressed as a. percentage of Patients treated with androgens, was as follows :
Table 2.
Hirsutism was the most embarrassing and unpleasant side- effect for the patient and was often very considerable. Not only was there an intense growth of long coarse hair on the cheeks, upper lip, chin, and general beard area but also the excess hair on the body, arms, and legs was considerable. It is the former, however, to which the patient naturally objects. Patients developing hypertrichosis were all recommended to shave daily, preferably with an electric razor, and not to resort to depilatory agents such as wax, pumice, emery paper, electrolysis, and plucking, all of which methods tend to cause infection of the skin follicles and unsightliness noticed by all, whereas, with a well-kept shaven face, the fact that the patient shaves is known only to herself and her husband. One patient with particularly heavy hirsuties controlled her appearance extremely well by rubbing off the hair from time to time with the pith side of lemon peel.
Voice changes probably depend on the normal pitch of the patient’s voice. They developed early in some and very late in others. The voice may be very deep and gruff or husky, or may crack and change like that of a pubescent boy. Patients sometimes commented on their inability to sing in the same register, which may drop an octave or so. This was rarely a source of dismay but is irreversible.
Enlargement of the clitoris was sometimes considerable although not invariable. The glans was often enlarged to 0-8-1-0 cm. in diameter, appeared congested, and protruded from the enlarged prepuce, whose sulcus contained smegma. The body of the clitoris was often 2—3 cm. long. Concomitant with this overgrowth of the clitoris there was increased vaginal secretion.
Acne developed only occasionally. The skin was more greasy in most of the patients in whom acne developed, and the acne then usually affected the face, chin, back, chest, and shoulders and probably depended partly on personal cleanliness.
All the above-mentioned virilising side-effects were seen with all the androgens used, occurring almost as much with methyl androstenediol as with testosterone and its esters. Although methyl androstenediol is said to be not so virilising, it certainly was so in the dosage of 100 mg. daily for a long time.
Increase of libido was an important side-effect (cf. Foss 1951) ; 14 patients (24-6%) -admitted to a considerable increase of libido, but the actual incidence was probably higher. It may be associated with hypertrophy of the clitoris. Patients commented on constant irritability and sensitivity, but in some cases where examination revealed an enlarged clitoris an increase in libido was denied.
Results
Survival.—-55-5% of the 106 patients were still alive 6 months after starting hormone treatment, 31-2% after 1 year, 11-6% after 2 years, 4-7% after 3 years and after 4 years, and 0-94% after 5 years. The average age of these 106 women at the beginning of therapy was 55 years and the mean survival for the group 48 weeks. The mean number of weeks lived in 5 years was 18-6% of the expected survival (Foss 1955). After the percentage expectancy of life in 5 years had been calculated for each case, statistical analysis gave the following results :
Table 3.
The differences between groups 1 and 2, and between groups 2 and 3 were not significant, but between groups 1 and 3. the expectancy of life was so increased as to be outside the range of random chance. If the survival figures from the onset of symptoms are used instead of from the start of hormone treatment, we find that the average age at onset of symptoms was 51-2 (range 26-76), and the average survival 58-5 months (range 10 months- 27 years. Dividing the patients into age-groups we get:
Table 4.
These figures indicate that the prognosis in women aged more than 60 at onset was worse than in younger women : 97-3% of the patients were still alive after 6 months from the onset of symptoms, 87-8% after 1 year, 62-2% after 2 years, 46-0% after 3 years, 34-0% after 4 years, 26-4% after 5 years, 5-6% after 10 years, and 2-8% after 15 years.
Improvement was observed in 31 patients (29%) : 1 aged less than 40, 19 aged 41-60, and 11 aged 61 or more. In 13 of the 14 women treated with oestrogen the first symptom had started after the menopause. (Estrogen therapy was started 4-27 years (mean 15 years) after the onset of the menopause. 7 of these patients had not undergone operation, because the cancer was too advanced when they were first seen. The average age of the 14 women was 63, and their average survival was 32 months : a woman on 0-3 mg. q.i.d. survived 59 months with quiescent bone métastasés ; another on 0-3 mg. q.i.d. survived 291/2 months with temporary regression of soft-tissue métastasés ; a woman on 5-0 mg. q.i.d. of more survived 17 months with quiescent bone métastasés ; and 11 women on 5-0 mg. q.i.d. or more survived 31-7 months (mean) with good regression of soft-tissue métastasés.
Only 17 patients (30%) of those treated with androgens showed some improvement ; their average age was 49 and average survival 17-25 months. Of these 17 women 14 (82%) had their first symptom before the menopause (1 underwent hysterectomy 11 years before the first symptom), and 4 had had neither surgery nor radiotherapy. The results of treatment were as follows :
Table 5.
The different types of métastasés responded as follows :
Of the 55 patients with bone métastasés 35 were treated with androgens, and of 31 complaining of severe pain 11 were relieved, including the 4 showing improvement radiographically. 6 patients had pathological fractures (4 during treatment) in which there had been no clinical improvement. In one the fractures were present on starting androgen treatment and subsequently healed : in another a fracture occurred during the early phase of treatment but later healed, together with radiographic improvement in the other bone métastasés. After oestrogen treatment none of the remaining 20 patients who had bone métastasés showed clinical or radiographic improvement, and pathological fractures occurred in 4 during the course of therapy.
In addition, less objective evidence of improvement was obtained by assessing the effect of treatment on the following :
(1) Pain or distress, which occurred in 80 patients, 20 of whom obtained relief.
(2) Serial haemoglobin levels. Before the start of treatment anæmia was found in 49 patients, 24 of whom had radiographically proved bone métastasés. There was virtually no improvement with treatment.
(3) The E.S.B. was raised initially in 68 patients and fell during treatment in 17.
(4) The body-weight increased in 23 (13 given androgen, 10 given oestrogen), decreased in 53, and did not change in 30.
(5) Blood-alkaline phosphatase : no definite correlationwas found between clinical improvement or extension or regression of bone métastasés judged by radiographic appear-ances. In one or two cases with bone secondaries which improved with testosterone, where the initial pre-treatment levels were high, subsequent values were much reduced. On the other hand, some cases treated with oestrogen showed similar changes in phosphatase with progressive deterioration in osseous métastasés.
Discussion
A therapeutic trial of this kind can only help to asses the effect of alteration of the steroid hormone balance of the host and makes no allowance for the specific malignancy of the cancer. Unfortunately the complete Broders’s grading of each case was unknown. In a few cases no details of the histology were available and many years had elapsed since the original operation.
In the preliminary trials it soon became apparent that only the old postmenopausal women_with soft-tissue métastasés might benefit.
The survival of any one patient with breast cancer of this advanced stage of severity must obviously depend on numerous factors—the patient’s age, the distribution and type of métastasés, the grade of malignancy, the j hormone dependence (whether of cestrogens or possibly of pituitary hormone), and other general health factors.
Spontaneous regression has been reported by Gordon Taylor, but malignancy may lie dormant for years. In the present series, however, the 106 patients had very advanced stage-xv métastasés. Even so, some patients were in reasonable general condition, whereas others were almost moribund. No cases were selected except that those with clinical evidence of considerable livermétastasés were excluded.
Comparison is necessary with the survival of patients I with untreated advanced mammary cancer :
Wade (1946) reported 27 untreated patients with breast cancer of whom 24 were in stage III. The average durationof life was 8 months from first being seen, and all were d in 2 years.
Greenwood (1926) found a mean survival of 38-3 months from the onset of symptoms in 651 cases ; Wyard (1925)found the mean duration of life to be 37 months in Sil cases; I and Daland (1927) quotes 40-5 months in 100 cases. Wade has collected these numbers together and found that the I average duration of life in untreated cases was 38-55 months I in 777 patients, and that the duration of life increased from the age-group 25-34 to that of 65-74 and then decreasedagain.
According to Wade (1946) X-irradiation alone or together with surgery raised the survival time by 21-5% at one year, 22-5% at 2 years, 10% at 3 yearn, 10-5% at 4 years, and 9-5% at 5 years over the survival-rates of untreated cases. In her series the mean survival of 27 patients with advanced stage-III untreated cancer whea I first seen was only 8 months ; hence 12 months’ survival in the present series of advanced stage-IV cancer after treatment with hormones is possibly significant.All her patients had died in 2 years, whereas 11-6% of tne present series were alive at 2 years and 4-7% at 4 years.
As regards surgical treatment alone, Lewison et al. (1953) claim a 5-year survival-rate of 43-2%,regardless of the type of operation. These were presumably selecte cases, inoperable stage-IV cancer being almost certainly excluded.
In the present series most of the patients had reeceived standard surgical and X-ray treatment in audition to hormones, and the mean duration of life'was months from the first symptom. There was a 5-year survivrate of 26-4% from onset. Although superficially this compares unfavourably, it must be remembered that my series consists solely of stage-iv advanced cancer.
The age at onset seems to influence duration of life. In patients who developed cancer of the breast when over 60 the percentage expected survial in 5 years was 20% less than in patients aged less than 40, and 30% less than in patients aged 41-60. A similar prognosis was reported by Burdick and Chanatry (1954).
In the present series of advanced stage-IV cancer the percentage expectancy of life in 5 years was significantly greater after hormone treatment in the patients aged more than 60 than in the younger patients. Although breast cancer is more rapidly fatal if it begins at the age of more than 60,. there is more chance of prolonging survival in this age-group by hormone therapy when métastasés are widespread.
Assessment of improvement was made as objective as possible. Temporary regression of métastasés in glands, photographically recorded improvement in skin lesions, and radiographic evidence of bone healing were accepted as beneficial effects of hormone therapy. 5 cases were regarded as improved, because pain of bone secondaries was diminished, although no radiographic improvement was seen during long periods of survival in moderate health. Only 4 of 35 patients with bone secondaries treated with androgen showed radiographic evidence of bone healing. This figure of 11% compares unfavourably with 30% claimed by Hallberg et al. (1953), who also claim improvement in general condition and palliation in 85%. In the present series objective improvement was seen in 30% of those treated with androgen—a very similar figure to 27% found in a shorter observation period by Segaloff et al. (1951) in 48 cases treated with testosterone propionate 100 mg. thrice weekly.
Subjective improvement was noted in a few other patients in this series treated with androgens, but it was doubtless partly due to the anabolic effect—increased appetite and temporary euphoria.
For some old patients who present with fungating lesions too advanced for surgery or X-irradiation cestro- gens are a useful means of palliation. Healing of ulceration in some cases may last months.
Huseby (1954), who treated with oestrogens 100 patients with advanced mammary cancer, reported that those in whom the response was favourable had an average survival of 29 months from the start of therapy, and that 21 of these were more than 5 years past the menopause (natural or therapeutic). The response of those with bone lesions was poor. Onjy 1 of his 5 patients who had not reached the menopause improved. There was major regression of pulmonary métastasés in 11. These figures are similar to my finding in 14 patients who improved with œstrogens. Their average age was 63 (range 53-76) and average survival 32 months from the start of therapy.
The considerable virilisation caused by big dosages of androgens is perhaps not so very important. Shaving with an electric razor will maintain a presentable appearance. Methyl androstenediol, which is claimed to be less virilising, has nothing to recommend it over testosterone. Patients treated with 100 mg. of methyl androstenediol daily for some months developed almost as much hirsutism as those given testosterone esters ; moreover methyl androstenediol causes pain when it is given in crystalline suspension, and in several cases it caused an abscess at the site of injection. Segaloff et al. (1952) observed similar results and achieved regression of the lesions in only 2 of 24 cases treated solely with methyl androstenediol.
The choice of therapy was not dictated at first by a knowledge of oestrogen dependence of a particular cancer. Usually old postmenopausal women with soft-tissue métastasés were treated with oestrogens, and the younger patients or those with widespread bone métastasés were given androgens. Subsequent analysis of the 17 patients who responded to androgens, however, showed that 82% had had their first symptom of cancer before the menopause whereas of the 14 who improved with oestrogens 93% had had their first symptom after the menopause. This may possibly be a guide to selection for hormone therapy, and in the androgen-treated group with premeuopausal onset this may indirectly indicate oestrogen dependence ; or, conversely, the postmenopausal onset sugests that not even stroma-cell oestrogens were responsible stimulants to their cancer (Sommers et al. 1953). In these old patients, in whom pituitary activity is not inhibited by endogenous oestrogens, there may be excessive production of somatotrophin and prolactin as well as gonadotrophin, and the beneficial effect of oestrogen therapy may be due to inhibition of these hormones.
Iff is quite clear that there is a type of patient with breast carcinoma which is temporarily responsive to treatment with androgens or by oophorectomy or by adrenalectomy, and that the essential common factor is oestrogen dependence (Pearson et al. 1954). Unfortunately there is no way of detecting oestrogen dependence, and only recently have chemical methods of estimating oestrogens in urine been elaborated, but the method is laborious (Brown 1955) and is not yet readily adaptable to clinical use.
In the present series the patients were so widely riddled with metastases that one could not hope for much benefit; hence an improvement-rate of about 30% for both oestrogens and androgens is possibly satisfactory.
As Asher (1955) aptly remarked : “ Prolongation of life is not the only aim of treatment; it should be a tolerable life. It is better to be wholly alive for one month than half alive for two.” The latter part of his statement referred to cases after endocrine surgery, which, like hormone therapy at present, suffers from an inability to discriminate between cancer cases with and without hormone dependence.
In the present series the side-effects and possible upsets and the prolonged course of oft repeated intramuscular injections were a high price to pay by patients for no improvement; perhaps it would have been better if they had been allowed to spend their remaining life under increasing doses of morphine. So much, however, depends on each patient’s temperament. One can only remember the gratitude and faith of those who were fortunate in their response to treatment, some of whom enjoyed many months’ freedom from pain and freedom from odour, and lived reasonably normal lives. Such results, even though restricted to about a third of those treated, are worth while.
Summary
106 women with stage-iv mammary carcinoma were treated with either androgen or oestrogen and kept under observation until death.
Improvement was observed in only 29% of cases. Of the 17 women who improved with androgen 14 had had their first symptom before the menopause ; and of 14 patients who improved with oestrogen 13 had had their first symptom after the menopause.
The average age of those patients who received benefit from oestrogen was 63 and from androgen 49. The prognosis was better for the older patients treated with oestrogen, the average survival for these being 32 months, compared with 17-25 months for those who improved with androgen.
The possible significance of oestrogen dependence in the younger group and pituitary-hormone dependence in the older postmenopausal patients is discussed.
My thanks are due to Sir Charles Dodds, who persuaded me to join in the British Empire Cancer Campaign oestrogen trial; the British Empire Cancer Campaign for a grant for three years for laboratory assistance ; Prof. Milnes Walker for permission to treat patients in the department of radiotherapy at the Bristol General Hospital; the physicians and surgeons who allowed me to treat their patients; Miss G. M. Jeffree, my technician; Dr. C. H. G. Price and Dr. A. L. Taylor for advice and help ; and Miss J. E. Nixon, of the records department of the department of radiotherapy, for much secretarial help. The androgens were provided by Ciba Laboratories, Organon Laboratories Ltd., and British Drug Houses Ltd. %
References
Adair, F. E., Herrmann, J. B. (1946) Ann. Surg. 123, 1023.
Asher, R. (1955) Proc. R. Soc. Med. 48, 376.
Beatson, G. T. (1896) Lancet, ii, 162.
Boyd, S. (1900) Brit. med. J. ii, 1161.
Brown, J. B. (1955) Lancet, i, 320.
Burdick, D., Chanatry, F. (1954) Cancer, 7, 47.
Cade, S. (1955) Proc. ii. Soc. Med. 48, 549.
Cutler, M., Schlemenson, M. (1948) J. Amer. med. Ass. 138, 187. Daland, E. M. (1927) Surg. Gynec. Obstet. 44, 264.
Dodds, E. C. (1955) Proc. R. Soc. Med. 48, 550.
Douglas, M. (1952) Brit. J. Cancer, 6, 32.
Foss, G. L. (i951) Lancet, i, 667.
— (1955) Proc. R. Soc. Med. 48, 551.
Greenwood, M. (1926) Rep. publ. Tilth med. Subj., Land. no. 33. Haddow, A. (1935) Nature, Loud. 136, 868.
Hallberg, O., Nohrman, B„ Sylven, B. (1953) Acta radiol., Stockh. 39, 161.
Herrmann, J. B., Adair, F. E. (1946) J. clin. Endocrin. 6, 769. Huggins, C., Bergenstal, D. M. (1952) Cancer Res. 12, 134.
— Dao, T. L. Y. (1952) Ann. Surg. 136, 595.
Huseby, R. A. (1954) Amer. Surg. 20, 112.
Lett, H. (1905a) Med.-Chir. Trans. 88, 147.
— (1905b) See Lancet, i, 227.
Lewison, E. F., Trimble, F. H., Griffith, P. C. (1953) J. Amer. med. Ass. 153, 905.
Loeser, A. A. (1938) Brit. med. J. ii, 319.
— (1939) C.R. Soc. franc. Qynéc. 9, 77.
— (1940) Brit. med. J. i, 479.
Lowcnhaupt, E., Steinbach, H. L. (1949) Surg. Gynec. Obstet. 88, 291. McWhirter, R„ Paterson, E., Thurgar, C. J. L., Walker, J. Z., Windeyer, B. W., et al. (1944) Proc. R. Soc. Med. 37, 731. Morris, H. (1902) Brit. med. J. ii, 1293.
Pearson, O. H., West, C. D., Hollander, V. P. (1954) J. Amer. med. Ass. 154, 234.
Prudente, A. (1948) Surg. Gynec. Obstet. 80, 575.
Schinzinger (1889) Verh. dtsch. Ges. Chir. 18, 28, quoted by Halber- staedter, L„ Hochman, A. (1946) J. Amer. med. Ass. 131, 810. Segaloff, A., Gordon, D., Horwitt, B. N., Schlosser, J. V., Murison, P. J. (1951) Cancer, 4, 319.
— — — — — (1952) Ibid. 5, 271.
Sommers, S. C., Teloh, H. A., Goldman, G. (1953) Arch. Surg. 67, 916.
Thompson, A. (1902) Brit. med. J. ii, 1538.
Wade, P. (1946) Brit. J. Radiol. 19, 272.
Wyard, S. (1925) Brit. med. J. i, 206.