Eur I Cancer Clm Oneal. Vol. 21. No. 10, pp. 1153-1158. 1985.
`Printed in Great Britain.
`
`0277-5579/85”.00+0.00
`© 1985 Pergamon Press Ltd.
`
`Aminoglutethimide in Advanced Breast Cancer:
`Clinical Results of a French Multicenter
`
`Randomized Trial Comparing 500 mg and
`l g/ day
`
`J. BONNETERRE,‘ H. COPPENS,T L. MAURIAC, M. METZ, J. ROUESSE, J. P. ARMAND, P. FARGEOT,
`M. MATHIEU, M. TUBIANA and P. CAPPELAERE‘
`
`‘Centre Oscar Lambret, Rue F. Combemale, BP 307, 59020 Lille Cédex, France and TLaboratoire Ciba Geig'y,
`2-4 Rue Lionel Terray, 92506 Rueil Malmaison, France
`
`Abstract— We have conducted a mullicenter randomized clinical trial comparing in
`advanced post-menopausal breast cancer patients 500 mg vs 1 g A 6/day. The
`hydrocortisone dose was 40 mg/day in both groups. One hundred and seventy
`patients have been randomized; 161 were evaluable for tolerability, 149 for
`effectiveness. Response rates were similar in both groups, 19 and 24% respectively
`for the 500 mg and 1 g groups. Na difference was observed according to tumor site.
`Duration of response was the same in both groups (14 months), as was mean time to
`response (about 3 months). Survival (studies in 125 patients) was similar in both
`groups (responders and non-responders). No response could be obtained with 1 g
`after relapse 0r failure with 500 mg (n = 17). Tolerability was good in 91% of the
`500 mg group patients and 78% of the 1 g group patients (P < 0.03). It was poor in 4
`and 15% respectively (P < 0.03). Side-effects were the same in both groups but less
`frequent and less severe in the 500 mg group; however,
`these patients more
`frequently had ‘moon face’.
`
`
`
`INTRODUCTION
`
`AMINOGLUTETHIMIDE (AG) has been widely used
`as a treatment for advanced breast cancer in post-
`menopausal women [1, 2]. Its use has also been
`proposed as an adjuvant treatment [3]. The AC
`dosage was previously 1 g/day with 40 mg
`hydrocortisone. In 1980 it was shown that, in vivo,
`lower concentrations of AG had the same activity
`as l g on estrone and estradiol plasma levels [4];
`these results were in agreement with those
`obtained in in vitro studies [5] which have shown
`aromatase to be ten times more sensitive to AG
`
`than desmolase; the preliminary results of in vivo
`hormonal studies have been largely confirmed
`
`Accepted 23 May 1985.
`List of participants and affiliations: Dr Chauvergne, Dr
`Mauriac: Fondation Bergonié, Bordeaux; Dr Fargeot: Centre
`G.F. Leclerc, Dijon; Dr Cappelaere, Dr Bonneterre, Dr
`Homer: Centre 0. Lambret, Lille: Dr Gary-hobo, Dr Mathieu:
`Centre P. Iamarque. Montpellier; Dr Metz: Centre A. Vautrin,
`Nancy; Dr Clavel, Dr Tubiana: Centre R. Huguenin, St Cloud;
`Dr Armand: Centre C. Regaud, Toulouse; Dr Rouesse, Dr
`Sevin: Institut G. Roussy, Villejuii.
`
`[6, 7]. These studies suggested that objective
`clinical responses could be obtained with lower
`doses of AG.
`We
`conducted a multicenter
`
`randomized
`
`trial comparing the effectiveness and
`clinical
`tolerability of 500 mg AG vs 1 g AG plus 40 mg
`hydrocortisone. The last patient was taken into
`this trial in August 1983.
`
`MATERIALS AND METHODS
`
`~
`Patient selection '
`One hundred and seventy spontaneously or
`radiation-induced menopausal women with
`proven metastatic breast carcinoma and measur-
`able progressive disease were selected for study.
`Patients with central nervous system involvement
`or hepatic metastasis were excluded from the
`study. Chemotherapy and/or other additive
`hormone therapy had to be discontinued at least 1
`month prior to entry into this trial. Among these
`170 patients, one was found to be ineligible
`(pulmonary cancer), eight were not evaluable
`
`1153
`
`AstraZeneca Exhibit 2144 p. 1
`InnoPharma Licensing LLC v. AstraZeneca AB IPR2017-00905
`
`

`

`1154
`
`J. Bonneterre et al.
`
`(neither for tolerability nor for effectiveness), 161
`were evaluable for
`tolerability and 149 were
`evaluable for effectiveness. The reasons for not
`
`treatment of too short a
`being evaluable were:
`duration, non-evaluable lesions, patients drop-
`ping out and treatment discontinuation because
`of toxicity. Most of these patients had been heavily
`pretreated: 68% of the 149 evaluable patients had
`already received at least one hormone therapy
`and/or chemotherapy.
`Patients were
`stratified according to the
`receptor status. All patients were either receptor-
`positive (R+) (estradiol and/or progesterone) or
`receptor-unknown (RP). Receptor-negative pat-
`ients were not eligible. Receptor assays were
`performed by the dextran-coated charcoal method.
`Positivity criteria were those accepted in each
`participating center.
`simultaneous
`No patient
`received
`therapy at an evaluable site.
`
`radio-
`
`Patient evaluation and response criteria
`Patients were evaluated after 15, 45 and 90 days,
`and then every 3 months. Response criteria were
`those of the UICC. Objective responses (whether
`complete or partial) and stable disease had to
`persist at
`least 3 months to be taken into
`consideration. Complete response (CR) means
`disappearance of all lesions. Partial response (PR)
`means regression of more than 50% of the product
`of two perpendicular diameters of all measurable
`lesions; stable disease (SD) was defined as a
`decrease of less than 50% or an increase of less than
`
`25% in all the lesions, and progressive disease (PD)
`as a progression of more than 25% in the lesions
`
`and/ or appearance of new lesions. Responses in
`bone metastases were considered only in the case
`of clear-cut radiological evidence of bone healing:
`regression of more than 50% in sclerosing
`metastases and evidence of calcification of more
`
`than 50% of the area of lytic lesions. Regression of
`bone pain, without evidence of bone healing, as
`defined, or with exaggerated calcification were
`not considered as a response, even in the case of
`regression of other metastatic sites.
`The duration of PR and SD was the interval
`
`the treatment and
`between the beginning of
`evidence of relapse. Duration of CR was the
`interval between the time that disappearance of
`the lesions was first observed and the first evidence
`
`of relapse.
`Subjective effectiveness (regression of pain,
`dyspnea) was classified as $0 in the case of
`worsening or no change in symptoms, 81 in the
`case of great improvement and S2 in the case of
`complete disappearance of all symptoms.
`The patient’s performance status (P5) was
`evaluated according to the WHO criteria (PS 0/1
`
`refers to patients in good general condition, able
`to work, PS 2 to patients unable to work but able
`to take care of themselves and PS 3/4 to patients
`confined to bed more than 50% of the daytime).
`
`Treatment protocol
`Patients were randomly allocated to treatment
`with either 500 mg or 1 g AG/day. All patients
`took 500 mg for the first 2 weeks (250 mg twice a
`day); from day 15 the patients of the 1 g group
`received 250 mg in the morning, 250 mg in the
`afternoon and 500 mg at bedtime. Hydrocortisone
`dosage was constant at 40 mg/day throughout the
`treatment, 10 mg at 8 a.m., 10 mg at 4 p.m. and
`20 mg at bedtime in both groups.
`
`Statistical analysis
`All survival curves were calculated by the
`method of actuarial survival analysis and the
`difference between pairs of curves was tested for
`statistical significance using the log rank test.
`
`RESULTS
`
`One hundred and sixty-one patients were
`evaluated for tolerability and 149 for effectiveness;
`clinical characteristics of these patients appear in
`Table 1.
`
`the
`of
`characteristics
`1. Clinical
`Table
`evaluable patients in the two groups
`
`
`
` 500 mg 1000 mg
`
`No. of patients
`
`73
`
`76
`
`Age (yr)'
`
`59.8 :1: 10.2
`
`59.7 i 10.6
`
`Relapse-free
`interval (months)‘
`
`Previous hormone
`therapy
`
`R+T
`R?1‘
`
`Metastatic sites:
`
`51.4 i 57.4
`
`45.9 i 35
`
`44%
`
`56%
`
`25
`48
`
`29
`47
`
`22
`15
`lymph nodes
`38
`39
`bone
`34
`26
`skin
`
`
`21lungs 14
`
`‘m :1:
`
`1 SD.
`
`TR+, steroid receptor (oestradiol and/or progesterone)-
`positive; RP,
`steroid receptor
`(oestradiol
`and
`progesterone) unknown.
`
`Clinical response according to AG dosage and
`metastatic site (Table 2)
`The overall response rates were 19% in the
`500 mg group and 24% in the lg group.
`If
`stabilization was included, the response rates were
`the same (58%) for both groups.
`'
`
`AstraZeneca Exhibit 2144 p. 2
`
`

`

`Aminoglutethimide in Advanced Breast Cancer
`
`1155
`
`Table 2. Response rates according to AC
`dosage, metastatic site and receptor status
`
`Table 3. Response rates according to
`previous hormone therapy
`
`AG regimen
`
`CR — PR CR — PR - SD
`
`Overall
`response
`
`500 mg
`1000 mg
`
`(14) 19%
`(18) 24%
`
`(42) 58%
`(44) 58%
`
`Bone
`
`Skin
`
`Lymph-
`node
`
`Lung
`
`R+
`
`R?
`
`‘ P < 0.05.
`
`500 mg
`1000 mg
`
`500 mg
`1000 mg
`
`500 mg
`1000 mg
`
`500 mg
`1000 mg
`
`500 mg
`1000 mg
`
`500 mg
`1000 mg
`
`(4) 10%
`(3)
`8%
`
`(8) 31%
`(5) 15%
`
`(5) 33%
`(8) 36%
`
`(4) 19%
`(7) 50%
`
`(7) 28%
`(8) 28%
`
`(29) 74%‘
`(20) 53%
`
`(14) 54%
`(19) 56%
`
`(10) 67%
`(16) 73%
`
`(13) 62%
`(10) 71%
`
`(15) 60%
`(19) 66%
`
`(7) 15%
`(10) 21%
`
`(27) 56%
`(25) 53%
`
`Patients
`
`n
`
`Response
`to AC
`PR — CR
`
`With previous hormone
`(15) 15%
`101
`therapy
`(8) 29%
`28
`Responders
`(6) 10%
`58
`Non-responders
`(1)
`7%
`15
`Not evaluable
`(12) 25%
`48
`Without previous
`hormone therapy
`
`Table 4. Response rates according to the
`No. of previous different hormone therapies
`and to AC dosage
`
`No. of responders (%)
`No. of
`different
`500 mg
`1 g
`hormone therapies
`(n = 48)
`(n = 53)
`
`
`3/20 (15%)
`4/18 (22%)
`1
`6/25 (24%)
`5/22 (23%)
`2
`
`
`
`1/8 (12%) 0/823 (0%)
`
`The difference in the response rates between the
`two groups was not statistically significant. When
`stable disease was included, the response rate for
`bone metastases was better in the 500 mg group.
`
`Clinical response according to AG dosage and
`receptor status (Table 2)
`Response rates were not different in R+ and R?
`patients, irrespective of whether they received 500
`mg or 1 g. Similarly, response rates were not
`different
`in group I and group 11 patients,
`irrespective of whether they were R+ or R?
`
`Mean time to response according to AC dosage
`The mean time before a response was observed
`was 3 months for both groups (2.8 :l: 1.2 months in
`the 500 mg group and 3.4 i 1.4 months in the 1 g
`group). In all cases the reSponse was observed
`within 6 months of the start of the study.
`
`Response rate according to performance status
`(PS)
`
`The response rate was similar, whatever the
`performance status of the patients at entry into the
`trial. Responses rates were 25% (n = 55), 14% (n =
`49) and 22% (n = 18), respeCtively, for PS 0/1 , PS 2
`and PS 3/4.
`
`hormone therapy (25%) and in patients who had
`responded to a previous hormone therapy with at
`least a minor response (29%). Response rates
`according to the number of different previous
`hormone therapies appear in Table 4. When the
`500 mg and the lg group patients are taken
`together,
`the response rates are 18% after one
`hormone therapy (n = 38), 23% after two (n = 47)
`and 6% after three or more different hormone
`
`therapies (n = 16).
`
`Subjective response
`Disappearance or improvement of subjective
`symptoms was noted in 68% of patients in both
`groups.
`
`Response to lg after failure with 500 mg AG
`In 17 patients randomly allocated to the 500 mg
`group and who failed to respond, a trial with l g
`was initiated; no response was observed.
`
`Duration of response according to AC dosage
`Median duration of response was 14 months in
`both groups. Median duration of stabiliration
`was 16 months in the 500 mg group and 10
`months in the 1 g group.
`
`Response rate according to previous hormone
`therapy (Tables 3 and 4)
`_
`Most of the patients had previously received
`hormone therapy or simultaneous hormone and
`chemotherapy which had most frequently been
`ineffective. The response rate was higher in
`patients who had never
`received previoUs
`
`Survival according to type of response and AG
`dosage
`Survival was studied according to the type of
`response in 125 patients, whatever the AG dosage
`(Fig.1). Considering both groups together, no
`difference was noted between patients who
`responded and those who were
`stabilized.
`
`AstraZeneca Exhibit 2144 p. 3
`
`

`

`1156
`
`]. Bonneterre et 31.
`
`However, a highly significant difference was
`found between these two groups and the non-
`responders (P < 0.001).
`As no difference in survival was noted between
`
`responders and patients who were only stabilized,
`they were considered together to compare survival
`in the 500 mg and 1 g groups. No difference was
`seen between the two groups (Fig. 2).
`
`The side-effects are listed in Table 6. Although
`lethargy and skin rash were much less frequent in
`the 500 mg group, the difference did not reach
`statistical significance. Conversely, ‘moon face’
`was more frequent in the patients of 500 mg group
`(P < 0.02). Four percent of the patients in the
`500 mg group and 7% in the 1 g group had to have
`AG stopped because of toxicity.
`
`Tolerability
`Tolerability was evaluated by the patient as
`good, fair or poor (Table 5). Tolerability was poor
`in 4% of the patients in the 500 mg group and 15%
`in the lg group (P < 0.03); it was good in 91 and
`78% (P < 0.03) of the two groups respectively.
`Tolerability was thus better in the 500 mg group.
`
`DISCUSSION
`
`Our results show that 500 mg AG has the same
`therapeutic activity with a better tolerance than
`1 g/day. No difference was observed between the
`two groups in reSponse rate, duration of response,
`mean time to complete or partial response, or
`survival. Furthermore, no response was seen with
`
`iOO
`
`0|0
`
`
`
`°/.survival
`
`
`
`6.
`
`A 27
`No. of patients
`at the beginning 0 45
`of each intervaL ‘ 53
`
`25
`40
`26
`
`is
`Months
`9
`I?
`3
`
`24
`
`30
`
`7
`I0
`2
`
`I
`0
`
`Fig. 1. Survival of all patients (whatever the AG dose) according to the type of response. Patients experiencing
`no change survived for as long as those who responded to the treatment.
`
`iOO
`
`‘
`
`x
`
`C
`
`\ NC-PRZACR 500 mg
`\.
`NC-PR-CR I000 mg
`
`
`
`6
`
`I~_~‘
`PD loco mg " PD 500 mg
`
`IS
`Months
`
`24
`
`30
`
`50
`
`"/asurvival
`
`No. of patients
`at the beginning
`of each interval
`
`A 35
`:
`. 26
`
`32
`
`,
`
`I
`
`28
`22
`7
`
`I6
`lg
`0
`
`ii
`g
`
`I
`2
`
`Fig. 2. Survival of patients according to the type of response (CR, PR and NC or PD) and AG dose. No
`difference in survival was noted between the l g and the 500 mg groups.
`
`AstraZeneca Exhibit 2144 p. 4
`
`

`

`Aminoglutethimide in Advanced Breast Cancer
`
`1157
`
`Table 5. Tolerability
`
`Good
`
`Fair
`
`Poor
`
`500mg
`(n = 78)
`
`(71)91%
`'l
`
`(4)5%
`
`4%
`
`(3)
`O
`
`(65) 78%
`
`(6) 7%
`
`(12) 15%
`
`1000 mg
`(n = 83)
`
`“ P < 0.03.
`
`Table 6. Side-effects
`
`500mg
`(n=78)
`
`9%
`(7)
`4%
`(3)
`3%
`(2)
`3%
`(2)
`(9) 12%
`(2)
`3%
`(1)
`1%
`(2)
`3%
`(2)
`3%
`(l)
`1%
`(3)
`4%
`
`1000 mg
`(n=83)
`
`(14) 17%
`(9) 11%
`(3). 4%
`(5)
`6%
`(l)
`1%
`(0)
`0%
`(1)
`1%
`(3)
`4%
`(2)
`2%
`(1)
`1%
`(7)
`8%
`
`"
`
`Drowsiness
`Skin rash
`Vertigo
`Ataxia
`Moon face
`Cramps
`Weight increase
`Hyperthermia
`Hypotension
`Hypertension
`Digestive symptoms
`
`'P < 0.02.
`
`l g when 500 mg was ineffective or after relapse
`with 500 mg.
`In this study, response rates in both the 500 mg
`and the l g groups were low compared with the
`published results [1,2]. This is probably due to
`the fact that most of the patients had been heavily
`pretreated, especially with hormone therapy
`(most of the patients had received several types of
`hormone therapy and chemotherapy). Further-
`more, the first hormone therapy was not evaluable
`or was ineffective in 72% of
`the cases;
`the
`
`likelihood of response to AG was therefore very
`low in these patients. The response rates in
`women who had already responded to hormone
`therapy and then relapsed and the response rates
`of those without any previous hormonal treat-
`ment are in keeping with published reports [8, 9].
`Our
`results in bone metastases were very
`disappointing, since Smith et al. [10] and Lipton
`et al. [1 1] had published objective response rates
`in 35% of 31 patients and in 33% of 27 patients.
`These results were much better
`than those
`
`obtained with tamoxifen. Response criteria are
`very difficult
`to assess and ours were strictly
`defined. The poor results that we obtained could
`be due to the fact that half of our patients had
`sclerosing or lytic and sclerosing metastases, the
`regression of which was particularly difficult to
`study. However,
`this
`cannot be
`the only
`explanation since the response rate of lytic lesions
`(eventually associated with sclerosing metastases
`but separately evaluated) was poor, too. Another
`
`explanation could be that bone healing may only
`be observed after a long-term treatment
`[2].
`However, the response rate for bone metastases,
`including stable disease, was 74% in the 500 mg
`group and 53% in the l g group.
`The response rates were not dependent on
`patients’ performance status. This had already
`been reported by Gale [12]. As it is well tolerated,
`this drug may safely be given to patients in a poor
`general condition, in contrast to chemotherapy.
`To our knowledge, such results have not been
`reported for other hormone therapies.
`Tolerability was better in patients receiving
`500 mg. Drowsiness, skin rash and ataxia were
`more frequent in patients treated with 1 gAG, but
`the difference was not statistically significant.
`However, patients treated with 500 mg felt much
`better. It should be emphasized that in the patients
`of our 1 g group tolerability was better than that
`reported in previous studies.
`In Santen and
`Brodie’s experience [1], skin rash was observed in
`30% of the cases and drowsiness in 33%. The
`
`incidence of ‘moon face’ in the patients of the
`500 mg group in our study might be due to an
`overcompensation with hydrocortisone. It seems
`possible that with 500 mg AG/ day desmolase is
`moderately inhibited; 20-30 mg hydrocortisone
`would probably be enough in patients receiving
`500 mg AG/ day. The ‘moon face” regressed after
`reduction of the hydrocortisone dosage.
`To our knowledge, this study is the first which
`compares two dosages of AG in a randomized
`trial. Harris et al. [6] has shown that estrone and
`estradiol plasma levels were suppressed by 125 mg
`AG given twice a day and that a further increase in
`the dosage did not result in a greater suppression
`of these plasma levels. Vermeulen et al. [7] showed
`that 125 mg AG caused a slight but significant
`decrease in estrone and estradiol plasma levels; 1 g
`AG did not produce a greater decrease in estrone
`and estradiol
`levels than 250 mg AG (20-40%).
`According to Vermeulen et al. [7], treatment with
`150 mg, 250 mg and l g AG reduced aromatase
`activity to 33%, 20% and 5% of the basal values
`respectively. These hormonal
`studies could
`explain why no response was obtained with 1 g
`AG after failure with 500 mg; the lowering of the
`estrone and estradiol plasma levels was probably
`sufficient with 500 mg AG to allow a response.
`Cantwell et al. [13], using 250 mg AG and 40 mg
`hydrocortisone, obtained 6 objective responses in
`20 patients. An alternative to the combined
`treatment
`is
`a
`lower AG dosage without
`hydrocortisone. Stuart-Harris et al. [14] treated 65
`patients with low-dosage AG alone (62.5 and
`125 mg twice daily); these dosages were as effective
`in lowering plasma estrone and estradiol levels as
`the standard 1 g dose, with minimal adrenal
`
`AstraZeneca Exhibit 2144 p. 5
`
`

`

`1158
`
`]. Bonneterre et al.
`
`inhibition. The response rate was 19%; sur-
`prisingly, side-effects were numerous in that
`study (lethargy 31%; skin rash 15%). Eight percent
`of
`these patients had to discontinue their
`treatment because of side-effects; Stuart-Harris et
`
`that hydrocortisone might
`suggest
`[14]
`al.
`diminish AG-induced side-effects.
`From the statistical point of view, it is not easy
`to show that two treatment arms give similar
`
`results; the number of patients to be included is so
`high that such trials are very difficult to conduct.
`However, these results together with our failure to
`obtain a response with 1 g after relapse with
`500 mg and together with the results of clinical
`non-comparative studies of low AG dosages [14]
`suggest that 500 mg is as effective — and better
`tolerated -— as l g.
`
`REFERENCES
`
`1. Santen R], Brodie AMI-I. Suppression of estrogen production as treatment of breast
`carcinoma: pharmacological and clinical studies with aromatase inhibitors. C lin
`Oncol 1982, 1, 77-130.
`
`2. Santen RJ, Henderson IC. A Comprehensive Guide to the Therapeutic Use of
`Aminoglutethimide. Basel, Karger, 1982.
`3. Coombes RC, Chilvers C, Dowsett M et al. Aminoglutethimide as an adjuvant to
`surgery in poor-risk post~menopausal patients with breast cancer. The Royal Society of
`Medicine. International Congress and Symposium Series, 1982, No. 53, 37—39.
`4. Paridaens R, Vermeulen A, Heuson JC. The inhibiting influence of increasing doses of
`aminoglutethimide on adrenal function and oestrogen production. Preliminary data.
`In: Paesi FJA, ed. Aminoglutethimfde. Mechanism of Action and Clinical Results in
`Breast Cancer. Basel, Ciba Geigy, 1982, 72-82.
`5. Graves PE, Salhaniclt HA. Stereo-selective inhibition of aromatase by enantiomers of
`amino-glutethimide. Endocrinology 1979, 105, 52-67.
`6. Harris AL, Dowsett M, Smith IE, Jeffcoate SL. Endocrine effects of low dose
`aminoglutethimide alone in advanced pOStmenopausal breast cancer. Br J Cancer 1983,
`47, 621-627.
`7. Vermeulen A, Paridaens R, Heuson JC. Effects of aminoglutethimide on adrenal
`steroid secretion. Clin Endocrinol 1983, 19, 673-682.
`8. Harris AL, Powles T], Smith IE et al. Aminoglutethimide for the treatment of advanced
`postmenopausal breast cancer. Eur J Cancer Clin Oncol 1983, 19, 11-17.
`9. Powles T]. The role of aromatase inhibitors in breast cancer. Semin Oncol 1983, 10,
`20—24.
`
`10. Smith IE, Harris AL, Morgan M, Gazet JC, McKinna JA. Tamoxifen versus
`aminoglutethimide versus combined tamoxifen and aminoglutethimide in the
`treatment of advanced breast cancer. Cancer Res 1982, 42, 34305-34335.
`11. Lipton A, Harvey HA. Santen R] et al. Randomized trial of aminoglutethimide versus
`tamoxifen in metastatic breast cancer. Cancer Res 1982, 42, 34345-34365.
`12. Gale KE. Treatment of advanced breast cancer with aminoglutethimide: a 14 year
`experience. Cancer Res 1982, 42, 33895-33965.
`13. Cantwell BM], Sainsbury JRC, Harris AL et al. Low-dose aminoglutethimide (AG);
`phase II study in advanced post-menopausal breast cancer. 2nd European Conference
`on Clinical Oncology, Amsterdam, November 1983, No. 17-09, 185.
`14. Stuart-Harris R, Dowsett M, Bozek T et al. Low dose aminoglutethimide in treatment
`of advanced breast cancer. Lancet 1984, ii, 604-607.
`
`AstraZeneca Exhibit 2144 p. 6
`
`