Articles
Preoperative gefitinib versus gefitinib and anastrozole in postmenopausal patients with oestrogen-receptor positive and epidermal-growth-factor-receptor-positive primary breast cancer: a double-blind placebo-controlled phase II randomised trial Andreas Polychronis, H Dudley Sinnett, Dimitri Hadjiminas, Hemant Singhal, Janine L Mansi, Dharsha Shivapatham, Sami Shousha, Jie Jiang, David Peston, Nigel Barrett, David Vigushin, Ken Morrison, Emma Beresford, Simak Ali, Martin J Slade, R Charles Coombes
Summary Background Some oestrogen-receptor (ER) positive breast cancers express epidermal growth factor receptor (EGFR), but whether inhibition of EGFR can suppress proliferation of breast cancer cells and ER function is not known. Methods In a double-blind, placebo-controlled randomised trial of 56 postmenopausal patients with ER-positive and EGFR-positive primary breast cancer, 27 women were randomly assigned to the tyrosine-kinase inhibitor of EGFR gefitinib (250 mg given orally once a day) and the aromatase inhibitor anastrozole (1 mg given orally once a day), and 29 women to gefitinib (250 mg given orally once a day) and placebo of identical appearance to anastrozole given orally once a day, all given for 4–6 weeks before surgery. Primary outcome was inhibition of tumour-cell proliferation, as measured by Ki67 antigen labelling index. Secondary outcomes were reduction in EGFR phosphorylation at Tyr 845, reduction in ER phosphorylation at Ser 118, tumour size, and toxic effects. Analyses were by intention to treat. Findings Patients assigned gefitinib and anastrozole had a greater reduction from pretreatment values in proliferation-related Ki67 labelling index than did those assigned gefitinib alone (mean % reduction 98·0 [95% CI 96·1–98·9] vs 92·4 [85·1–96·1]; difference between groups 5·6% [5·1–6·0], p=0·0054). Tumour size was reduced by 30–99% (partial response) in 14 of 28 patients assigned gefitinib and anastrozole and in 12 of 22 assigned gefitinib, as assessed by ultrasonography. Reduction in phosphorylation of ER at Ser 118 was similar for both groups. Treatment was well tolerated and much the same for both groups. Interpretation Single-agent gefitinib and gefitinib combined with anastrozole are well-tolerated and effective treatments for reducing the size of breast tumours and levels of ER phosphorylation when given as neoadjuvant therapy.
Introduction The proliferation of breast-cancer is frequently dependent on oestrogen, and most human breast cancer expresses oestrogen receptor (ER).1 However, trials2 in which presurgical (ie, neoadjuvant) systemic treatment was given to keep surgery to a minimum have shown that only about half of primary breast cancers respond to endocrine therapy, which frequently takes 3–6 months. For this reason, many physicians use cytotoxic chemotherapy, which is associated with substantial morbidity.3 Several mechanisms have been suggested for the reduced effectiveness of endocrine therapy. Epidermal growth factor receptor (EGFR), which is commonly expressed in breast cancer, is associated with a poor prognosis and failure to respond to endocrine therapy.4 2500 (48%) of 5232 breast tumours were EGFR positive in 40 different studies,5 depending on the method of detection used. Furthermore, immunohistochemical analyses have shown that 30% of ER-positive tumours were positive for EGFR, and in some series this proportion was as high as 51–77%.6–8 http://oncology.thelancet.com Vol 6 June, 2005
Activation of EGFR results in tyrosine autophosphorylation at several sites, including Tyr residue 845 in the activation loop of the receptor kinase domain.9,10 These events lead to the recruitment and phosphorylation of several intracellular substrates and to the binding of docking and adaptor molecules to specific phosphotyrosine sites on receptor molecules.11 EGFR therefore signals for survival of breast-cancer cells, which might attenuate the effects of oestrogen withdrawal induced by endocrine treatment.12 A key phosphorylation site in ER, Ser 118, is located in the ligand-independent transcription-activation function AF1.13 Ser 118 is phosphorylated via stimulation of mitogen-activated protein kinase (MAPK) by epidermal growth factor,14,15 resulting in ER activation in the absence of oestrogen.16 Furthermore, inhibitors of MAPK signalling can inhibit the phosphorylation of Ser 118 by epidermal growth factor,15 as can gefitinib (Ali S, unpublished data). However, Ser 118 is also phosphorylated on oestrogen binding through a pathway independent of MAPK, by recruitment of TFIIH-
Lancet Oncol 2005; 6: 383–91 Published online May 17, 2005 DOI: 10.1016/S1470-2045(05) 70176-5 Cancer Research UK Laboratories, Department of Cancer Medicine, Imperial College, London, UK (A Polychronis MRCP, J Jiang MD, M J Slade PhD, D Vigushin FRCP, S Ali PhD, Prof R C Coombes F MedSci); AstraZeneca UK, Luton, UK (K Morrison PhD, E Beresford MSc); Department of Surgery, St Mary’s Hospital, London, UK (D Hadjiminas FRCS); Department of Surgery, Northwick Park Hospital, London, UK (H Singhal FRCSEd); Department of Oncology, St George’s Hospital, London, UK (J L Mansi FRCP); and Department of Medical Oncology (D Shivapatham MSc), Department of Histopathology (S Shousha FRCPath, D Peston FIBMS), Department of Radiology (N Barrett FRCR), and Department of Breast Surgery, Charing Cross Hospital, London, UK (H D Sinnett FRCS). Correspondence to: Prof R Charles Coombes, Department of Cancer Medicine, Imperial College, Hammersmith Hospital Campus, 8th Floor, MRC Cyclotron Building, Du Cane Road, London W12 0NN, UK
[email protected]
383
Articles
studies are needed to define the relation between response and EGFR copy number in breast cancer, however amplification of EGFR has been recorded in only 0·8% of 2197 breast cancers.24
573 with primary breast cancer screened 225 excluded* 62 premenopausal 158 carcinoma in situ 55 had previous cancer 348 eligible at first screening 234 excluded* 89 ER negative 127 EGFR negative 37 premenopausal 114 eligible at second screening 57 excluded 20 out of catchment area 37 refused treatment 57 enrolled
1 withdrew consent
56 randomised
29 assigned gefitinib
29 analysed
27 assigned gefitinib and anastrazole
27 analysed
Figure 1: Trial profile *Some patients were excluded for more than one of these reasons.
associated protein kinase CDK7.17 Details of ER and EGFR crosstalk in breast cancer have been summarised by Ali and Coombes.18 We therefore postulated that inhibition of EGFR by tyrosine-kinase inhibition and of oestrogen biosynthesis by aromatase inhibition might result in substantial suppression of ER phosphorylation and therefore cellular proliferation. Other roles of EGFR suggest that it is a target for breast-cancer treatment, since ligand binding can lead to cell proliferation, motility, improved survival, invasion, and angiogenesis.19,20 Gefitinib, a 4-anilinoquinazolone,21 is an orally active, well tolerated inhibitor of EGFR, and at a dose of 250 mg a day leads to objective responses in lung cancer. Baselga22 has reviewed the role of gefitinib in the treatment of lung cancer. Moreover, a study23 of patients with metastatic colorectal cancer found that EGFR copy number correlated with response to antiEGFR treatment.23 Thus, whereas in lung cancer response correlates with the presence of specific mutations in some patients, in colorectal cancer and breast cancer such mutations are rarely found. Further 384
Methods Trial design and patient selection We did a multicentre, double-blind, placebo-controlled randomised phase II trial of 56 postmenopausal patients with ER-positive and EGFR-positive primary breast cancer. 29 patients were randomly assigned to 250 mg gefitinib and 1 mg oral anastrozole (Arimidex®, AstraZeneca, UK) given once daily, and 27 women to 250 mg gefitinib (Iressa®, AstraZeneca, UK) and placebo, both given orally once a day. Treatment was given for 4–6 weeks or until disease progression, unacceptable toxic effects, or withdrawal of consent. Definitive surgery was planned for the end of treatment. The placebo was identical in appearance to anastrozole. The primary endpoint of the trial was tumour-cell proliferation as measured by the change in Ki67 antigen labelling index. This proliferation-related antigen has been validated previously.25 Secondary endpoints were: reduction in EGFR phosphorylation at Tyr 845; reduction in ER phosphorylation at Ser 118; tumour size; and toxic effects. Tumour size was assessed clinically and by ultrasonography. Clinical size measurements were done by the same physician at every recruiting centre; ultrasonography was also used because it is regarded as more sensitive than mammography and palpation,26,27 and equal to MRI.28 Analyses were by intention to treat. The study was thought of, designed, and coordinated by The Department of Cancer Medicine, Imperial College, London, UK, and done with the help of the West of London Cancer Research Network, UK. Recruitment of patients was done at four hospitals in London, UK: Charing Cross; St Mary’s; St George’s; and Northwick Park hospital. The ethics review boards of every hospital reviewed the protocol. Patients gave written informed consent in accordance with the Declaration of Helsinki. Patients were eligible if they were postmenopausal (aged 50 years or older with amenorrhoea for more than 2 years), had histologically diagnosed ER-positive and EGFR-positive breast cancer. All patients had surgically resectable disease and carcinomas larger than 1 cm in the longest diameter as measured by ultrasonography or 2 cm as measured clinically. Ultrasonography before initial biopsy and after treatment was done by the same operator in every centre, and was centrally reviewed by NB. Patients had not received any previous treatment for breast cancer, and had no evidence of metastases as assessed by bone, liver, and chest imaging. All patients had adequate haematological function (neutrophil count 1500106 cells per L and platelets 140109/L), renal function (serum creatinine 1·5 times upper limit of http://oncology.thelancet.com Vol 6 June, 2005
Articles
normal), and liver function (bilirubin 1·25 times upper limit of normal and other liver-function tests 2·5 times upper limit of normal) at randomisation. A computer-generated random list of unique numbers was used to assign treatment. This information was placed in an individual unique-identifier-labelled envelope, which was sealed for the duration of the trial. One copy was stored at every clinical site, at the sponsor company, and at a third-party (the Scottish Poisons Information bureau) for emergencies. All individuals responsible for generating the list played no other part in the management, analysis, or reporting of this study. The unique identification numbers were allocated sequentially to patients at every centre by the designated site pharmacist. Once the study prescription with the unique number on it was signed by a physician, the patient took it to the pharmacist responsible for the study, who opened the next envelope in sequence describing the treatment allocation. Participating patients, medical staff, ancillary medical staff giving the treatments and assessing the welfare of the patient during the trial were blinded as to treatment assignment from the start of the randomisation period until the last patient had completed treatment and the data was gathered from the site. At no point in the study was there any indication to necessitate unblinding of any patients to treatment.
Analyses Core biopsy samples were assessed for Ki67 labelling, ER, progesterone receptor (PR), ERBB2, EGFR, phosphorylated MAPK, phosphorylated ER at Ser 118, and phosphorylated EGFR at Tyr 845, all of which were done at the start of the study. All patients had clinical examinations and biochemical testing once a week; clinical and ultrasonography findings were recorded in accordance with Response Evaluation Criteria In Solid Tumors (RECIST, partial=30–99% reduction, stable=0–29% reduction).29 Trial treatment was discontinued in all patients at the end of the trial, and was followed by surgery, adjuvant chemotherapy, and radiotherapy as indicated. All patients assigned treatment were assessed for adverse events. Patients analysed for primary and secondary endpoints were those who had evaluable biopsy samples at both baseline and end of treatment. Tumour response was assessed for patients with tumour assessments at baseline and end of treatment. Immunohistochemistry was done with 4-m sections from formalin-fixed, paraffin-embedded blocks after antigen retrieval by microwaving. Antibodies for ER (clone 6S11, Novacastra Laboratories, UK), PR (clone PR88, Menarini, UK), ERBB2 (A0485, DAKO, UK), Ki67 (Novocastra), phosphorylated EGFR at Tyr 845 (Cell signaling 2231, New England Biolabs, UK), and phosphorylated MAPK (Cell signaling 4376, New England Biolabs), and phosphorylated ER at Ser 118 (Cell signaling 2511, New England Biolabs) were used. http://oncology.thelancet.com Vol 6 June, 2005
Gefitinib (n=29) Age (years) Mean (SD) 64·7 (8·0) Histology Infiltrating ductal carcinoma 25 Infiltrating lobular carcinoma 4 Size of primary tumour on ultrasonography (cm) Mean (SD) 2·9 (1·6) 1–2 11 2–4 13 4 5 Size of primary tumour on clinical analysis (cm) 1–2 5 2–4 15 4 9 Tumour status Grade 1 1 Grade 2–3 28 Ki67 labelling index (%)* 0–29 10 30–69 17 70–100 0 ER status 26 3 0 PR status 14 / 7 8 ERBB2 status ERBB2 6 ERBB2 1 ERBB2 or – 22 Nodal status N1 4 Past hormone-replacement treatment Yes 8 Performance status WHO 0 27 WHO 1 2 Ethnic origin White 25 Black 0 Other 4
Gefitinib and anastrozole (n=27) 63·4 (8·8) 23 4 2·6 (2·4) 15 10 2 0 24 3 4 23 13 12 1 23 2 2 12 7 8 3 0 24 3 6 23 4 22 3 2
*Excludes two patients assigned gefitinib alone and one patient assigned gefitinib and anastrozle who had insufficient tissue for immunohistochemical analyses.
Table 1: Baseline characteristics
Antibody 2511 has been validated in our laboratory (Sarwar N, unpublished data) and by others.30 EGFR status was established by use of an antibody (cell signaling 2232) directed against the intracellular domain (New England Biolabs) and one against the external domain (Zymed antiEGFR [31G7], Zymed Laboratories, UK); microwaving and enzyme digestion, respectively, were used for antigen retrieval. The two antibodies to EGFR have been extensively validated in our laboratory. Competition with molar excess of the immunogen peptides eliminated staining on formalin-fixed paraffinembedded breast tissue (data not shown). Further confirmative evidence of antibody specificity was derived by use of RNA interference followed by 385
Articles
Ki67 (% cells stained)* Number of patients with baseline 0 Mean before treatment (95% CI) Mean after treatment (95% CI) Mean % reduction (95% CI) Tumour size reduction on ultrasonography (mm) Not done Partial response (30–99%) Stable disease (0–29%) Mean size before treatment (95% CI) Mean size after treatment (95% CI) Mean % reduction Tumour size reduction on clinical analysis (mm) Not palpable at baseline Complete response (100%) Partial response (30–99%) Stable disease (0–29%) P-Y845 EGFR (score)* Number of patients with baseline 0 Before treatment (95% CI) After treatment (95% CI) Number with reduction to 0 P-S 118 ER (score)* Number of patients with baseline 0 Before treatment (95% CI) After treatment (95% CI) Number with reduction to 0 P-MAPK (score)* Number of patients with baseline 0 Before treatment (95% CI) After treatment (95% CI) Number with reduction to 0
Gefitinib (n=29)
Gefitinib and anastrozole (n=27)
27 27·6 (19·9–38·3) 2·1 (1·0–4·4) 92·4 (85·1–96·1)
26 26·6 (19·2–37·0) 0·5 (0·3–1·1) 98·0 (96·1–98·9)
1 14 14 28·9 (22·7–35·2) 19·2 (15·2–23·2) 31·9 (41·5–22·3)
5 12 10 26·1 (16·5–35·8) 19·6 (12·9–26·2) 29·7 (37·1–22·3)
3 5 18 3
2 7 11 7
26 2·07 (1·70–2·45) 0·37 (0·06–0·68) 20
21 1·58 (1·16–1·99) 0·27 (0–0·54) 16
26 2·70 (2·44–2·97) 1·30 (0·83–1·76) 9
26 2·73 (2·49–2·97) 1·04 (0·65–1·43) 9
24 0·93 (0·77–1·08) 0·41 (0·18–0·63) 14
22 1·0 (0·77–1·23) 0·31 (0·06–0·56) 18
*Excludes two patients assigned gefitinib alone and one patient assigned gefitinib and anastrozole who had insufficient tissue for immunohistochemical analyses.
Table 2: Effects of treatment
immunocytochemistry. Western blotting of four breastcancer cell lines showed the presence of a 170 kDa band in known EGFR-positive cell lines (data not shown). Amount of EGFR, phosphorylated EGFR at Tyr 845, phosphorylated MAPK, ER, PR, ERBB2, and phosphorylated ER at Ser 118 were recorded independently by AP and SS, who were blinded as to treatment allocation. Samples were scored as negative (score 0) or positive (score + to +++)31 on the basis of the proportion of stained cells (nuclear for ER and PR staining; membranous, cytoplasmic, or both for EGFR, P-Y845 EGFR, ERBB2, and P-MAPK staining). For Ki67 labelling index, a total of 400 cells were counted from three consecutive high-power magnifications.
Statistical analysis 26 patients were needed in each treatment group to provide 90% power to detect a standardised difference of 1 (treatment difference/SD) between groups at the 5% significance level. On the assumption of a log normal distribution, Ki67 values were log transformed before analysis, and all means are therefore geometric means. Log(Ki67post) and log(Ki67pre) have been used to denote the mean of the logs of the three post-treatment 386
Gefitinib Gefitinib and (n=29) anastrozole (n=27) Number of patients with adverse events* Eye disorders Eye pruritus Keratoconjunctivitis sicca Gastrointestinal disorders Diarrhoea Dry mouth Nausea Loose stools Skin disorders Rash Dry skin Pruritus Dermatitis acneform Pruritic rash Vascular disorders Hot flush Pulmonary disorders
28 4 3 2 22 9 9 3 6 25 14 6 6 5 1 2 2 0
26 4 1 2 18 5 8 5 7 20 12 5 1 6 3 6 6 0
*Adverse events seen in 10% of patients in either group. Some patients had more than one adverse event in every category. Data are number of patients.
Table 3: Treatment-related adverse events
and pretreatment values respectively. 0·1 was added to every untransformed Ki67 value to avoid the mathematical anomaly that arises because the log of zero is minus infinity. As a consequence of the assumption of a lognormal distribution, log(Ki67post–logKi67pre), or log(Ki67post/Ki67pre), is also normally distributed. This formula gives the proportional change, and as a result mean log proportional changes and CI can be calculated and displayed on their original scale by back transformation. The change in log(Ki67) was compared between groups by use of the t test, and the proportional change within groups was analysed with the paired t test. The validity of these assumptions was confirmed by comparison of means and CI with medians and 95% CI of untransformed values. The proportional reduction was calculated as one minus the proportional change. Changes in secondary outcomes from baseline to posttreatment were analysed between treatment groups with the Wilcoxon Mann-Whitney test and within treatment groups with the Wilcoxon signed rank test. The change in tumour size was expressed as the proportional change from baseline to post-treatment; changes in the other secondary outcomes were expressed as the absolute change from baseline to post-treatment. Associations between outcomes were investigated by use of the Spearman’s rank correlation coefficient. A comprehensive statistical-analysis plan was prepared before unblinding of the data. Statistical tests were two-sided with an level of 0·05. Statistical analyses were done with SAS software version 8.1 by EB. Data and statistical analyses were checked independently by Folcrum Ltd (Hemel Hempstead, UK) and Syne Qua Non Ltd (Norfolk, UK). http://oncology.thelancet.com Vol 6 June, 2005
Articles
Role of the funding source The sponsor of this trial, AstraZeneca, was responsible for drug provision, monitoring of the trial in accordance with the International Conference on Harmonisation— Good Clinical Practice, and reporting of all serious adverse events and other events to the appropriate authorities. An AstraZeneca monitor was responsible for data collection. The funding source had no involvement in the trial design; analysis and interpretation of data; or in the writing of the article. The corresponding author had full access to the data, and was responsible for the decision to submit for publication.
Results Figure 1 shows the trial profile. We recruited 57 patients from four hospitals in west London between April 24, 2003, and Nov 29, 2004. Assessment of the biological effects of treatment was possible for 53 patients; three
A
patients (two assigned gefitinib alone and one assigned gefitinib and anastrozole) had insufficient tissue for immunohistochemical analysis. Table 1 shows patients’ baseline characteristics. Duration of treatment was much the same for both groups (mean 34·6 days [SD 12·8] in the gefitinib group and 34·3 days [7·7] in the combination group), although three patients in each group had treatment for 43–79 days. Both groups showed a reduction in Ki67 in response to treatment (p0·0001 for both groups, table 2). Patients assigned gefitinib and anastrozole had a greater decrease in Ki67 labelling index than did those assigned gefitinib and placebo (p0·0001). (difference between groups 5·6% [95% CI 5·1–6·0], p=0·0054; table 2). Clinical assessment of tumour size correlated with that of assessment by ultrasonography before treatment (correlation coefficient 0·84) and after
Gefitinib alone
90 Partial response
Stable disease
Partial response
Stable disease
80 70 60 50 40 30 20
Ki67 labelling index
10 0
B
Gefitinib and anastrozole
90 80 70 60 50 40 30 20 10 0 Day 1
Day 28
Day 1
Day 28
Figure 2: Disease response as assessed by ultrasonography and Ki67 score Red lines are those patients whose tumours showed complete abolition of phosphorylated ER.
http://oncology.thelancet.com Vol 6 June, 2005
387
Articles
treatment (0·84, p0·0001), and ultrasonography measurements correlated with pathological tumour size (0·70, p0·0001). Clinically, 52 of 56 patients had a breast carcinoma of 2 cm or more at the start of treatment. Ultrasonography showed about 50% of patients had a partial response. As we expected after a short treatment period, no pathological complete responses were recorded. None of the patients progressed during treatment. Figure 2 shows the relation between disease response as assessed by ultrasonography and Ki67 score. Abolition of phosphorylated EGFR at Tyr 845 was much the same in both groups (table 2). Phosphorylated MAPK immunostaining was abolished in a greater proportion of patients allocated gefitinib and anastrozole than in those assigned gefitinib alone (p0·0001 for reduction to 0 for both groups), but groups did not differ significantly (p=0·172). Before treatment
4 weeks after treatment
A
B
C
Treatment did not affect total ER (data not shown), but phosphorylated ER at Ser 118 fell during treatment in both groups (p0·0001). About a third of patients in each group had complete abolition of phosphorylated ER staining, mainly those who achieved a radiological partial response (figure 2). In patients assigned gefitinib and anastrozole who had stable disease on ultrasonography, one of ten showed complete abolition of phosphorylated ER compared with six of 12 who had a partial response (p=0·06, figure 2). Figure 3 shows examples of immunostaining results. Scores for phosphorylated EGFR and phosphorylated MAPK did not correlate with radiological response to treatment in either group. 12 of 19 patients with PR-positive tumours who were allocated gefitinib and anastrozole showed a reduction in PR staining after treatment compared with six of 21 allocated gefitinib and placebo. Two of three patients with ERBB2-positive tumours who were assigned gefitinib and anastrozole showed abolition of ERBB2 staining after treatment compared with four of seven in the gefitinib-alone group. Histological grade of breast carcinomas were identical before and after treatment in all but two patients who had a reduction. Table 3 shows adverse events. Gefitinib was well tolerated as a single agent and in combination with anastrozole. Mild WHO grade 1–2 adverse events seen in both groups are consistent with the known safety profiles of gefitinib and anastrozole. Treatment was interrupted for 3–11 days in three patients in the gefitinib-alone group because of transient grade 1 diarrhoea in two patients and grade 1 skin toxicity and abdominal symptoms in one patient. The addition of anastrozole did not exacerbate the adverse effects of gefitinib. WHO grade 3 or 4 treatment-related adverse events were seen in two patients assigned gefitinib and placebo (dry skin and elevated alanine transaminase, and dermatitis), and in one patient assigned gefitinib and anastrozole (skin rash). No pulmonary toxic effects associated with gefitinib were noted. Other events, not thought to be treatment-related, were myocardial infarction (n=1), cerebrovascular accident (n=1), and haemoptysis (n=1), all of which occurred in those assigned gefitinib and placebo.
Discussion D
Figure 3: Immunohistochemical analyses in a patient assigned gefitinib and anastrozole (A) Ki67 labelling index. (B) Y845 P-EGFR. (C) S118 P-ER. (D) P-MAPK (magnification200).
388
We have shown that gefitinib, either alone or in combination with anastrozole, substantially reduced tumour-cell proliferation (as assessed by Ki67 antigen labelling index) and tumour size in previously untreated patients with ER-positive and EGFR-positive breast cancer, and that combination treatment was more effective than was gefitinib alone. These effects occurred within 4 weeks of the start of treatment in most patients. However, our findings suggest that gefitinib alone is a useful treatment in primary breast cancer, since it also rapidly reduced tumour size. Although tumour size http://oncology.thelancet.com Vol 6 June, 2005
Articles
decreased in both groups, as measured clinically and radiologically, the reductions were not as extensive as the Ki67 reduction, possibly because of the short treatment period. In a validation study, we found no difference in Ki67 labelling index between the initial biopsy sample and the repeat definitive biopsy sample in patients with breast cancer who had not received treatment (Coombes RC, unpublished). Dowsett and colleagues25 showed that anastrozole alone inhibited Ki67 labelling index by 76% after 2 weeks’ treatment. Our findings suggest that gefitinib alone has similar effects after 4 weeks of treatment. These researchers also noted that the greater inhibition of Ki67 antigen by anastrozole compared with tamoxifen correlated with improved disease-free interval recorded in the Anastrozole or Tamoxifen Alone or in Combination trial. Our study suggests that gefitinib has a role in the adjuvant setting, and that a strategy of EGFR inhibition could have been applicable to 114 of 348 (33%) of postmenopausal patients screened. However, more detailed and longer studies are needed before such recommendations can be made. We found that treatment substantially reduced phosphorylation of EGFR at Tyr 845 and of MAPK, and that gefitinib is thus an effective inhibitor of EGFR activity. This finding has several explanations. EGFR is expressed by many breast-cancer cells,32 and EGFR inhibitors cause tumour regression in breast-cancer cell lines positive for EGFR.33 Gefitinib and other EGFR inhibitors have striking effects on the cell cycle and on apoptosis in vitro. Experiments with monoclonal antibody 225, which acts against EGFR, showed G1 arrest as a result of enhanced P27 and hypophosphorylation of retinoblastoma protein,34,35 leading to reduced activity of cyclin-dependent kinases 2, 4, and 6 and of cyclin D1 and D3. Cells also showed increased FAS ligand, with downstream enhancement of caspases 8 and 3.36 Such FAS-dependent apoptosis is consistent with the rapid reduction in tumour size seen in our study and in patients with lung cancer who are responsive to gefitinib.37 Gefitinib can inhibit ERBB2 signalling in breastcancer cells in vitro, even at concentrations that do not suppress ERBB2 tyrosine-kinase activity.38 ERBB2 preferentially heterodimerises with EGFR, ERBB3, and ERBB4. Gefitinib seems to sequester ERBB3, ERBB2, and EGFR in an inactive, unphosphorylated heterodimer configuration, resulting in suppression of heregulininduced receptor activation and proliferation in vitro.39 However, only a few patients in our study had tumours that expressed ERBB2, and this mechanism is an unlikely explanation for our results. Our study suggests that patients who have completely suppressed ER phosphorylation at Ser 118 are more likely to show an early radiological response than are those whose suppression is absent or incomplete. Findings from Ki67 labelling show an additional effect http://oncology.thelancet.com Vol 6 June, 2005
of combining gefitinib with anastrozole. Several in-vitro studies show that an antiEGFR strategy combined with antihormonal treatment could prevent the emergence of resistance to endocrine therapy, and that ERBB2 or MAPK inhibition could abrogate tamoxifen resistance.40 Other researchers41 have suggested that EGFR upregulation occurs with the onset of resistance to endocrine therapy, and that acquired tamoxifen resistance in human MCF-7 breast-cancer cells coincides with increased sensitivity to gefitinib.41 Since the start of our study, small groups of patients with metastatic breast cancer have been given gefitinib, and in these four studies,42–45 only one patient in every study showed clear evidence of response. However, an early clinical report46 in patients with tamoxifen-resistant breast cancer suggests that gefitinib reduces Ki67 in patients deriving clinical benefit. Because nearly all patients in our study responded to gefitinib with or without anastrozole by having a complete or almost complete (ie, 90%) abolition of Ki67, associating Ki67 reduction with that of reduction in EGFR and ER phosphorylation is difficult. However, patients in our study achieved a reduction in ER phosphorylation by EGFR inhibition, implying that ER is phosphorylated at Ser 118 in vivo—mainly by EGFR— and suggesting that dual inhibition of ER and EGFR prevents de-novo resistance to endocrine therapy. Pharmacokinetic analyses to investigate interactions between anastrozole and gefitinib showed no effect of anastrozole on gefitinib metabolism, and plasma concentrations of anastrozole and gefitinib were consistent with previous studies (data not shown). In conclusion, our findings that gefitinib alone and gefitinib and anastrozole reduced the Ki67 antigen labelling index suggest a role for EGFR inhibition in postmenopausal women with ER-positive and EGFRpositive primary breast cancer. Such a systemic approach to treatment might reduce the size of large primary breast cancers and enable patients to undergo conservative surgery. Neoadjuvant chemotherapy as primary systemic treatment is regarded the standard of care in the management of locally advanced breast cancer, irrespective of whether the tumour is ER positive on core biopsy sampling,47 and has been increasingly used in women with small tumours. Such treatment can result in local regression of the tumour or in a complete pathological response.48 In patients with ER-positive breast cancer, neoadjuvant hormonal therapy is an attractive alternative to cytotoxic chemotherapy because it is easy to give and avoids acute and late toxic effects of cytotoxic treatment.47 However, many clinicians are reluctant to use neoadjuvant hormonal treatment because of the slow rate of response. Gefitinib, combined with an aromatase inhibitor, might have a role in the neoadjuvant treatment of breast cancer by reducing the size of the tumour more rapidly. Studies are now being designed to assess this approach. 389
Articles
Contributors A Polychronis was co-investigator, and helped write the protocol and article, recruit patients, and did all clinical assessments and most of the immunostaining, with the help of J Jiang and D Peston. H D Sinnett, D Hadjiminas, H Singhal, D Vigushin, and J L Mansi (co-investigators) recruited and assessed patients at different sites. D Shivapatham was the data coordinator. S Shousha was the pathologist who reviewed all immunohistochemistry. N Barrett coordinated the radiology and reviewed all scans centrally. K Morrison did the overall data management and data analyses. E Beresford did the statistical analyses. S Ali and M Slade helped design the study and standardise laboratory methods. R C Coombes was the principal investigator and instigated the study, helped with the design, and with A Polychronis, wrote this paper. Conflict of interest A Polychronis has received an educational grant from AstraZeneca towards his PhD studies. K Morrison and E Beresford are AstraZeneca employees. The other authors declare no coflicts of interest. Acknowledgments This study was funded by AstraZeneca, UK (study code 1839i-0077). We thank The National Translational Cancer Research Network, The West London Cancer Research Network (T Daniels), Cancer Research UK, and The Breast Cancer Research Trust for support; J Lewis, J Grigono, R Cummins, and the nurses of the breast units at Charing Cross, St Mary’s, St George’s, and Northwick Park hospitals for helping to recruit patients; Annali Harris for help with data collection; Roger A’Hern for help with statistical analyses; Syne Qua Non Ltd and Folcrum Ltd; and Ms J Bliss for critical review. References 1 Ali S, Coombes RC. Estrogen receptor alpha in human breast cancer: occurrence and significance. J Mammary Gland Biol Neoplasia 2000; 5: 271–81. 2 Wong ZW, Ellis MJ. First-line endocrine treatment of breast cancer: aromatase inhibitor or antioestrogen? Br J Cancer 2004; 90: 20–25. 3 Mano MS, Awada A. Primary chemotherapy for breast cancer: the evidence and the future. Ann Oncol 2004; 15: 1161–71. 4 Arpino G, Green SJ, Allred DC, et al. HER-2 amplification, HER-1 expression, and tamoxifen response in estrogen receptor-positive metastatic breast cancer: a Southwest Oncology Group study. Clin Cancer Res 2004; 10: 5670–76. 5 Klijn JG, Berns PM, Schmitz PI, Foekens JA. The clinical significance of epidermal growth factor receptor (EGF-R) in human breast cancer: a review on 5232 patients. Endocr Rev 1992; 13: 3–17. 6 Bevilacqua P, Gasparini G, Dal Fior S, Corradi G. Immunocytochemical determination of epidermal growth factor receptor with monoclonal EGFR1 antibody in primary breast cancer patients. Oncology 1990; 47: 313–17. 7 Lefebvre MF, Garin E, Falette N, Saez S. Evaluation of epidermal growth factor receptor occupancy by EGF-like peptide in 55 breast and 42 non-breast tumour biopsies. Eur J Cancer 1990; 28: 182. 8 Wrba F, Reiner A, Ritzinger E, et al. Expression of epidermal growth factor receptors (EGFR) on breast carcinomas in relation to growth fractions, estrogen receptor status and morphological criteria. An immunohistochemical study. Pathol Res Pract 1988; 183: 25–29. 9 Hubbard SR, Wei L, Ellis L, Hendrickson WA. Crystal structure of the tyrosine kinase domain of the human insulin receptor. Nature 1994; 372: 746–54. 10 Cooper JA, Howell B. The when and how of Src regulation. Cell 1993; 73: 1051–54. 11 Schlessinger J. Cell signaling by receptor tyrosine kinases. Cell 2000; 103: 211–25. 12 Baselga J. Continuous administration of ZD1839 (Iressa), a novel oral epidermal growth factor receptor tyrosine kinase inhibitor (EGFR-TKI), in patients with five selected tumour types: evidence of activity and good tolerability. Proc Am Soc Clin Oncol 2000; 19: 686. 13 Ali S, Metzger D, Bornert JM, Chambon P. Modulation of transcriptional activation by ligand-dependent phosphorylation of the human oestrogen receptor A/B region. EMBO J 1993; 12: 1153–60.
390
14
15
16
17
18 19 20
21
22 23
24
25
26
27
28
29
30
31
32
33
34
35
Kato S, Endoh H, Masuhiro Y, et al. Activation of the estrogen receptor through phosphorylation by mitogen-activated protein kinase. Science 1995; 270: 1491–94. Chen D, Washbrook E, Sarwar N, et al. Phosphorylation of human estrogen receptor alpha at serine 118 by two distinct signal transduction pathways revealed by phosphorylation-specific antisera. Oncogene 2002; 21: 4921–31. Bunone G, Briand PA, Miksicek RJ, Picard D. Activation of the unliganded estrogen receptor by EGF involves the MAP kinase pathway and direct phosphorylation. Embo J 1996; 15: 2174–83. Chen D, Riedl T, Washbrook E, et al. Activation of estrogen receptor alpha by S118 phosphorylation involves a liganddependent interaction with TFIIH and participation of CDK7. Mol Cell 2000; 6: 127–37. Ali S, Coombes RC. Endocrine-responsive breast cancer and strategies for combating resistance. Nat Rev Cancer 2002; 2: 101–12. Yarden Y, Sliwkowski MX. Untangling the ErbB signalling network. Nat Rev Mol Cell Biol 2001; 2: 127–37. Mendelsohn J, Baselga J. Status of epidermal growth factor receptor antagonists in the biology and treatment of cancer. J Clin Oncol 2003; 21: 2787–99. Wakeling AE, Guy SP, Woodburn JR, et al. ZD1839 (Iressa): an orally active inhibitor of epidermal growth factor signaling with potential for cancer therapy. Cancer Res 2002; 62: 5749–54. Baselga J. The science of EGFR inhibition: a roadmap to improved outcomes? Signal 2004; 5: 4–8. Moroni M, Veronese S, Benvenuti S, et al. Gene copy number for epidermal growth factor receptor (EGFR) and clinical response to antiEGFR treatment in colorectal cancer: a cohort study. Lancet Oncol 2005; 6: 279–86. Al-Kuraya K, Schraml P, Torhorst J, et al. Prognostic relevance of gene amplifications and coamplifications in breast cancer. Cancer Res 2004; 64: 8534–40. Dowsett M, Smith IE, Ebbs SR, et al. Short-term changes in Ki-67 during neoadjuvant treatment of primary breast cancer with anastrozole or tamoxifen alone or combined correlate with recurrence-free survival. Clin Cancer Res 2005; 11: 951S–58S. Yeh E, Slanetz P, Kopans DB, et al. Prospective comparison of mammography, sonography, and MRI in patients undergoing neoadjuvant chemotherapy for palpable breast cancer. AJR Am J Roentgenol 2005; 184: 868–77. Berg WA, Gutierrez L, NessAiver MS, et al. Diagnostic accuracy of mammography, clinical examination, US, and MR imaging in preoperative assessment of breast cancer. Radiology 2004; 233: 830–49. Londero V, Bazzocchi M, Del Frate C, et al. Locally advanced breast cancer: comparison of mammography, sonography and MR imaging in evaluation of residual disease in women receiving neoadjuvant chemotherapy. Eur Radiol 2004; 14: 1371–79. Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to evaluate the response to treatment in solid tumors. European Organization for Research and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada. J Natl Cancer Inst 2000; 92: 205–16. Murphy LC, Niu Y, Snell L, Watson P. Phospho-serine-118 estrogen receptor-alpha expression is associated with better disease outcome in women treated with tamoxifen. Clin Cancer Res 2004; 10: 5902–06. McCarty KS Jr, Miller LS, Cox EB, et al. Estrogen receptor analyses. Correlation of biochemical and immunohistochemical methods using monoclonal antireceptor antibodies. Arch Pathol Lab Med 1985; 109: 716–21. Sainsbury JR, Farndon JR, Needham GK, et al. Epidermal-growthfactor receptor status as predictor of early recurrence of and death from breast cancer. Lancet 1987; 1: 1398–402. Ciardello F, Caputo R, Bianco R, et al. Antitumor effect and potentiation of cytotoxic drugs activity in human cancer cells by ZD-1839 (Iressa), an epidermal growth factor receptor-selective tyrosine kinase inhibitor. Clin Cancer Res 2000; 6: 2053–63. Wu X, Rubin M, Fan Z, et al. Involvement of p27KIP1 in G1 arrest mediated by an anti-epidermal growth factor receptor monoclonal antibody. Oncogene 1996; 12: 1397–403. Peng D, Fan Z, Lu Y, et al. Anti-epidermal growth factor receptor monoclonal antibody 225 up-regulates p27KIP1 and induces G1
http://oncology.thelancet.com Vol 6 June, 2005
Articles
36
37
38
39
40
41
arrest in prostatic cancer cell line DU145. Cancer Res 1996; 56: 3666–69. Chang GC, Hsu SL, Tsai JR, et al. Molecular mechanisms of ZD1839-induced G1-cell cycle arrest and apoptosis in human lung adenocarcinoma A549 cells. Biochem Pharmacol 2004; 68: 1453–64. Fukuoka M, Yano S, Giaccone G, et al. Multi-institutional randomized phase II trial of gefitinib for previously treated patients with advanced non-small-cell lung cancer. (The IDEAL 1 Trial). J Clin Oncol 2003; 21: 2237–46. Campiglio M, Locatelli A, Olgiati C, et al. Inhibition of proliferation and induction of apoptosis in breast cancer cells by the epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor ZD1839 (‘Iressa’) is independent of EGFR expression level. J Cell Physiol 2004; 198: 259–68. Anido J, Matar P, Albanell J, et al. ZD1839, a specific epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor, induces the formation of inactive EGFR/HER2 and EGFR/HER3 heterodimers and prevents heregulin signaling in HER2overexpressing breast cancer cells. Clin Cancer Res 2003; 9: 1274–83. Kurokawa H, Lenferink AE, Simpson JF, et al. Inhibition of HER2/neu (erbB-2) and mitogen-activated protein kinases enhances tamoxifen action against HER2-overexpressing, tamoxifen-resistant breast cancer cells. Cancer Res 2000; 60: 5887–94. Gee JM, Harper ME, Hutcheson IR, et al. The antiepidermal growth factor receptor agent gefitinib (ZD1839/Iressa) improves
http://oncology.thelancet.com Vol 6 June, 2005
42
43
44
45
46
47
48
antihormone response and prevents development of resistance in breast cancer in vitro. Endocrinology 2003; 144: 5105–17. Baselga J, Albanell J, Ruiz A, et al. Phase II and tumor pharmacodynamic study of gefinitib (ZD1839) in patients with advanced breast cancer. Proc Am Soc Clin Oncol 2003; 22: 7 (abstr). Robertson JF, Gutteridge E, Cheung KL, et al. Gefitinib (ZD1839) is active in acquired tamoxifen (TAM)-resistant oestrogen receptor (ER)-positive and ER-negative breast cancer: results from a phase II study. Proc Am Soc Clin Oncol 2003; 22: 7 (abstr). von Minckwitz G, Jonat W, Fasching P, et al. A multicentre phase II study on gefitinib in taxane- and anthracycline-pretreated metastatic breast cancer. Breast Cancer Res Treat 2005; 89: 165–72. Tan AR, Yang X, Hewitt SM, et al. Evaluation of biologic end points and pharmacokinetics in patients with metastatic breast cancer after treatment with erlotinib, an epidermal growth factor receptor tyrosine kinase inhibitor. J Clin Oncol 2004; 22: 3080–90. Gee JM, Gutteridge E, Robertson JF, et al. Biological markers during early treatment of tamoxifen-resistant breast cancer with gefitinib (Iressa). Breast Cancer Res Treat 2004; 88 (suppl 1): S32. Kaufmann M, von Minckwitz G, Smith R, et al. International expert panel on the use of primary (preoperative) systemic treatment of operable breast cancer: review and recommendations. J Clin Oncol 2003; 21: 2600–08. Fisher B, Brown A, Mamounas E, et al. Effect of preoperative chemotherapy on local-regional disease in women with operable breast cancer: findings from National Surgical Adjuvant Breast and Bowel Project B-18. J Clin Oncol 1997; 15: 2483–93.
391
