- Email: [email protected]
Clinical, pathological, proliferative and molecular responses associated with neoadjuvant aromatase inhibitor treatment in breast cancer
Journal of Steroid Biochemistry & Molecular Biology 118 (2010) 273–276
Contents lists available at ScienceDirect
Journal of Steroid Biochemistry and Molecular Biology journal homepage: www.elsevier.com/locate/jsbmb
Clinical, pathological, proliferative and molecular responses associated with neoadjuvant aromatase inhibitor treatment in breast cancer夽 W.R. Miller ∗ Breast Unit Research Group, University of Edinburgh, Western General Hospital, Edinburgh, United Kingdom
a r t i c l e
i n f o
Article history: Received 1 September 2009 Received in revised form 2 October 2009 Accepted 7 October 2009 Keywords: Aromatase inhibitors Response Breast cancer
a b s t r a c t Neoadjuvant treatment provides an exceptional setting in which to monitor clinical, pathological, proliferative and molecular responses to aromatase inhibitors. Sequential measurements of the primary tumour provide an accurate assessment of clinical changes and the relatively easy access to the tumour within the breast means that biopsies are available for histological and molecular measurements before and during treatment. Large randomised trials (P024 and IMPACT) together with informative nonrandomised studies have demonstrated clinical responses to third generation aromatase inhibitors in 40–70% of ER-positive tumours, rates generally significantly higher than observed with tamoxifen. Pathological responses in terms of reduced cellularity/increased fibrosis are also seen in 65–75% of cases. Whilst these are more often seen in clinically responding tumours, the correlation between clinical and pathological response is not absolute. A marked feature of treatment with third generation inhibitors is a reduction in cellular proliferation. Using Ki67 as a marker, this may be observed as early as 10–14 days into treatment. Reduction in proliferation with treatment may be seen in both clinically responding and non-responding tumours, although incidence and degree of effect are higher in responding cases. Aromatase inhibitor treatment frequently fails to reduce proliferation in tumours over-expressing HER-2. In terms of molecular events, aromatase inhibitor treatment is associated with changes in expression of genes classically associated with oestrogen regulation (KIAA0101, ZWINT, IRS1 and TFF1) and cell cycle progression, most notably mitotic phase proteins (CDC2, CCNB1 and CKS2). Changes occur both in clinically responding and non-responding tumours. Although expression of no individual gene correlates absolutely with response status, expression signatures can be produced which distinguish between responding and non-responding tumours. In terms of gene ontology, terms relating to macro-molecular biosynthesis, translation and structural components of ribosomes are significantly enriched. Finally, molecular signatures can be used to illustrate the relative homogeneity of responding tumours and the disparate nature of non-responding tumours suggesting multiple and diverse pathways associated with resistance. © 2009 Elsevier Ltd. All rights reserved.
1. Introduction Neoadjuvant treatment regimes in which aromatase inhibitors are given to patients with the primary tumour still within the breast provide excellent protocols by which to measure responses associated with therapy [1–3]. Firstly, clinical response may be accurately assessed by sequential measurement of the primary tumour. Additionally, because the cancer is invariably biop-
夽 Special Issue selected article from the IX International Aromatase Conference (Aromatase 2008) held at Shanghai, China, on October 13th–16th, 2008. ∗ Current address: 2 Stoneycroft Road, South Queensferry EH30 9HX, United Kingdom. E-mail address: [email protected]. 0960-0760/$ – see front matter © 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.jsbmb.2009.10.005
sied to confirm diagnosis and to determine oestrogen receptor (ER) status, tissue may be available to assess pathology, histology, proliferation and molecular phenotype/genotype and to relate these to clinical response. Because, in most protocols, patients come to definitive breast surgery after 3–4 months’ therapy, tumour is also available at this time and the effects of treatment may be measured on the same pathological and biological parameters. Experimental protocols have also been described in which additional biopsies have been taken after 10–14 days treatment so that early effects of therapy may be monitored and used to predict tumour response at 3–4 months [4,5]. The paper reviews data derived from the neoadjuvant use of aromatase inhibitors in terms of observations on clinical, pathological and proliferative responses. It also describes how these interrelate,
274
W.R. Miller / Journal of Steroid Biochemistry & Molecular Biology 118 (2010) 273–276
and may be used to understand the biological effects of aromatase inhibition in breast cancer and optimise patient management in those with the disease. 2. Clinical response Clinical response rates for third generation aromatase inhibitors varies between 40 and 75% [3]. Two major clinical trials (P024 and IMPACT) have randomised neoadjuvant use of aromatase inhibitors versus tamoxifen [6,7]. Benefits were greater for the aromatase inhibitors; in the P024 trial the difference was statistically significant (55% for letrozole versus 34% tamoxifen, p < 0.001) whereas that in IMPACT (anastrozole versus tamoxifen) did not reach statistical significance. 3. Pathological response Marked changes in tumour morphology and histology may be associated with aromatase inhibitor treatment. In oestrogen receptor-rich tumours effects on histological features include changes in cellularity, degree of fibrosis and histological grading. These changes in tumour morphology may be seen by 3–4 months as evidenced by a clear reduction in cellularity or increase in fibrosis [8,9]. Such pathological responses have been reported in 60–80% of tumours [10,11], although complete pathological responses are rarely seen within the 3–4 months timeframe [12]. Treatment may also reduce histological grade; mitotic score being the most commonly affected grading feature [11]. Although rates of clinical and pathological responses are similar and may be significantly associated, there is not exact concordance, and about a fifth of tumours display a discordant phenotype in which tumours either display a decrease in tumour volume without evidence of pathological changes or do not change in tumour size whilst displaying a decrease in cellularity [13]. 4. Proliferative response Marked decreases in tumour proliferation (as evidenced by reduced mitotic score/staining of Ki67) in ER+ve breast cancers have been observed following treatment with aromatase inhibitors [10,12]. These changes may be seen in about 80% of cases after 3 months’ treatment (with a median decrease of about 90%). As compared with neoadjuvant tamoxifen treatment the effects are generally more pronounced and seen in a greater proportion of cases [5,10]. However, reduced proliferation may be observed much earlier in treatment; for example, studies in which biopsies have been taken at 10–14 days of treatment show marked and significant decreases in the expression of Ki67 [13]. Although similar effects may be seen at these early time-points with tamoxifen, the effects are significantly greater with aromatase inhibitors [14]. Several studies have examined the patterns of changes in proliferation with time. The majority of tumours display decreases in proliferation at 10–14 days which are maintained or become more pronounced with treatment after 3 months. However, it is possible to observe in a minority of tumours other patterns in which (a) initial decrease in proliferation at 10–14 days is followed by recovery towards pre-treatment values at 3 months, (b) reduction in proliferation is seen at 3 months but not 10–14 days and (c) cases display little changes either at 10–14 days or 3 months [13]. Proliferative responses are positively correlated with both clinical and pathological response [13–15]. However, they are not equivalent and discordant phenotypes occur relatively frequently. For example, clinical response and cell cycle response to both tamoxifen and letrozole in the P024 trial were discordant in over one third of cases [15]. Similarly the degree of proliferative response
was not significantly different in cases clinically responding or resistant to anastrozole in the IMPACT trial [14]. However, it does appear that the relationship between proliferative response and pathological response is greater than that between proliferative response and clinical response [13], perhaps reflecting the closer association between two histological assessments over that between histology and tumour size. Interestingly, aromatase inhibitor treatment often fails to reduce proliferation in tumours over-expressing HER-2 [16–19]. 5. Molecular response Clear evidence of molecular responses has been elicited following treatment with aromatase inhibitors. For example, the progesterone receptor (PgR) which is classically regarded as a marker of oestrogenic activity is clearly reduced in 70–80% of breast cancers and, in about one half of cases, it may completely disappear [12,13]. Effects may be seen as early as 10–14 days and are consistent with the anti-oestrogenic effects of aromatase inhibitors. Loss of PgR expression may occur independently of pathological and clinical response [13,20]. Similar effects have been noted on the expression of other oestrogen regulated markers such as pS2 [20]. These effects may also be contrasted to those produced by neoadjuvant tamoxifen in which decreases in PgR and pS2 are much less frequent and, indeed, increases in expression may be often observed (reflecting the oestrogen agonist effects of this SERM). The use of microarray technology has great potential in identifying genes which are markers of molecular response and may be associated with/predict for clinical, pathological and proliferative response in breast cancers. There have been recent key publications [21–25]. For example, early changes in gene expression have been identified by comparing microarray analyses from paired tumour core biopsies taken before and after 14 days’ neoadjuvant treatment with letrozole [21–23]. These have confirmed that aromatase inhibitor treatment is associated with changes in expression of genes classically associated with oestrogen regulation such as KIAA0101, ZWINT, TFF1 and IRS1, and cell cycle progression most notably mitotic phase proteins such as CDC2, CCNB1 and CKS2. Although more frequent in clinically responding tumours, these changes also occur in clinically non-responsive tumours [24,25]. As a consequence, they are unlikely to be reliable predictors of clinical response. 6. Prediction of clinical response Because markers of oestrogen action and cell proliferation are likely to be imperfect predictors of clinical response, it is important to identify novel genes or expression profiles which have a greater utility. To do this, microarrays from pre- and 14-day treatment tumour biopsies have been compared in patients either responding or resistant to neoadjuvant letrozole treatment [22,24,25]. Expression of no single gene correlated absolutely with response status. However, expression signatures can be derived which distinguish between responding and non-responding tumours. This work has been published in a preliminary form [22,24] and is about to be published in full [25]. In brief, it has been possible to identify 205 covariables (69 baseline genes, 45 day 14 genes and 91 changes in gene expression between baseline and day 14) which are associated with clinical response status. In terms of molecular function, gene ontology analysis reveals ‘structural constituents of ribosomes’ is highly significantly enriched and, in terms of biological processes, terms such as macromolecule biosynthetic processes, cellular biosynthetic processes and translation are also significantly enriched. Complete linkage hierarchical clustering of cases on the basis of the covariables separates responders from non-responders
W.R. Miller / Journal of Steroid Biochemistry & Molecular Biology 118 (2010) 273–276
275
Fig. 1. Heatmap of the dissimilarity matrix between cases and corresponding dendrograms (top and left dendrograms are identical). Bar colours beneath the dendrogram code for response—blue for responding cases and purple non-responding. Colour tones in the dissimilarity matrix represent degree of similarity—red and orange indicate high similarity, yellow and white high dissimilarity. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of the article.)
into two distinct groups. It also illustrates that responding tumours comprise a relatively homogenous group clearly dissimilar genetically from the non-responding tumours (Fig. 1). 7. Markers of resistance Molecular studies may also be utilised to illustrate the relative heterogeneity of non-responding tumours [24,26]. For example, using classical markers of oestrogen regulation such as KAA0101, SERPINA3, IRS1, TFF3, TFF1 and proliferation such as CDC, CKS2, cyclin B1, TYMS, PCNA, it is possible to subdivide tumours resistant
to letrozole into three subgroups: (i) cases which are molecularly resistant both in terms of oestrogen markers and proliferation (a phenotype similar to the lack of gene expression changes seen in ER−ve tumours); (ii) oestrogen sensitive tumours which display a phenotype of a reduction in oestrogen regulated genes but no change or an increase in expression of cell cycle markers, and (iii) molecularly sensitive tumours displaying a reduction in both oestrogen and cell proliferation markers (see Fig. 2). These observations illustrate the disparate nature of non-responding tumours and suggest multiple and diverse pathways are associated with resistance.
Fig. 2. Changes in expression of oestrogen regulated and proliferation associated genes after 14 days adjuvant treatment with letrozole. Colour coding—green represents downregulation and red upregulation. Intensity of colour represents degree of change. Single column on left of figure represents changes in an ER−ve case; the other panels relate to 15 different clinically resistant ER+ve tumours—the left-hand panel illustrates a molecular resistant phenotype, the middle panel cases showing decreases in the expression of oestrogen regulated genes but not in cell cycle genes, and the right-hand panel molecular sensitivity in both oestrogen regulated and cell cycle associated gene expression. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of the article.)
276
W.R. Miller / Journal of Steroid Biochemistry & Molecular Biology 118 (2010) 273–276
8. Conclusions This review has illustrated how neoadjuvant treatment specifically using aromatase inhibitors presents a unique opportunity by which to assess the clinical, pathological, proliferative and molecular changes associated with this type of endocrine therapy in individual breast cancers. It is clear that whilst the different types of response are inter-related, there are relatively frequent discordant phenotypes where, for example, clinically responsive tumours fail to show pathological changes and clinically resistant tumours may show clear molecular responses. The understanding of the mechanisms behind these phenomena will undoubtedly help to rationalise treatment, to identify mechanisms of resistance and develop circumventing or synergistic treatment regimes; all of which should produce clinical benefits both in terms of higher response rates and more enduring tumour responses. Molecular phenotyping has also produced evidence of a relative homogeneity between tumours which are clinically responsive to aromatase inhibitors. Conversely, changes in expression of genes related to oestrogen signalling and proliferation can illustrate heterogeneity between resistant tumours. The implication of this heterogeneity is that multiple pathways and mechanisms are involved in resistance to aromatase inhibitors and that diverse strategies may be needed to circumvent such resistance. Genetic signatures are also being developed which can distinguish between responsive and resistant tumours. In future these may be used not only to predict cases in which to use aromatase inhibitors, but also to suggest alternative drug regimes which may be successful in resistant tumours. In summary, neoadjuvant therapy with aromatase inhibitors is not only producing clinical benefits in terms of reducing tumour size and allowing more conservative breast surgery, but it is providing fundamental knowledge on pathological, biological and molecular processes which will impact on the future use of these endocrine agents. References [1] M. Kaufmann, G.N. Hortobagyi, A. Goldhirsch, S. Scholl, A. Makris, P. Valagussa, J.U. Blohmer, W. Eiermann, R. Jackesz, W. Jonat, A. Lebeau, S. Loibl, W. Miller, S. Seeber, V. Semiglazov, R. Smith, R. Souchon, V. Stearns, M. Untch, G. von Minckwitz, Recommendations from an international expert panel on the use of neoadjuvant (primary) systemic treatment of operable breast cancer: an update, J. Clin. Oncol. 24 (12) (2006) 1940–1949. [2] J.M. Dixon, C.D. Love, C.O. Bellamy, D.A. Cameron, R.C. Leonard, H. Smith, W.R. Miller, Letrozole as primary medical therapy for locally advanced and large operable breast cancer, Breast Cancer Res. Treat. 66 (3) (2001) 191–199. [3] J.M. Dixon, J. Jackson, L. Renshaw, W.R. Miller, Neoadjuvant tamoxifen and aromatase inhibitors: comparisons and clinical outcomes, J. Steroid. Biochem. Mol. Biol. 86 (3–5) (2003) 295–299. [4] W.R. Miller, T.J. Anderson, D.B. Evans, A. Krause, G. Hampton, J.M. Dixon, An integrated view of aromatase and its inhibition, J. Steroid Biochem. Mol. Biol. 86 (2003) 413–421. [5] W.R. Miller, T.J. Anderson, S. White, A. Larionov, J. Murray, D. Evans, A. Krause, J.M. Dixon, Aromatase inhibitors: cellular and molecular effects, J. Steroid Biochem. Mol. Biol. 95 (2005) (2005) 83–89. [6] W. Eiermann, S. Paepke, J. Appfelstaedt, A. Llombart-Cussac, J. Eremin, J. Vinholes, L. Mauriac, M. Ellis, M. Lassus, H.A. Chaudri-Ross, M. Dugan, M. Borgs, Letrozole Neo-Adjuvant Breast Cancer Study Group, Preoperative treatment of postmenopausal breast cancer patients with letrozole: a randomized doubleblind multicenter study, Ann. Oncol. 12 (11) (2001) 1527–1532. [7] I.E. Smith, M. Dowsett, S.R. Ebbs, J.M. Dixon, A. Skene, J.U. Blohmer, S.E. Ashley, S. Francis, I. Boeddinghaus, G. Walsh, IMPACT Trialists Group, Neoadjuvant treatment of postmenopausal breast cancer with anastrozole, tamoxifen, or both
[8] [9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
in combination: the Immediate Preoperative Anastrozole, Tamoxifen, or Combined with Tamoxifen (IMPACT) multicenter double-blind randomized trial, J. Clin. Oncol. 23 (22) (2005) 5108–5116. J.M. Dixon, C.D.B. Love, L. Renshaw, et al., Lessons from the use of aromatase inhibitors in the neoadjuvant setting, Endocr. Relat. Cancer 6 (1999) 227–230. H. Sasano, T. Suzuki, T. Moriya, Pathology of breast cancer following neoadjuvant therapy, in: W.R. Miller, J.N. Ingle (Eds.), Endocrine Therapy in Breast Cancer, Marcel Dekker, New York, USA, 2002, pp. 213–222. W.R. Miller, J.M. Dixon, L. Macfarlane, et al., Pathological features of breast cancer response following neoadjuvant treatment with either letrozole or tamoxifen, Eur. J. Cancer 39 (4) (2002) 462–468. T.J. Anderson, J.M. Dixon, M. Stuart, et al., Effect of neoadjuvant treatment with anastrozole on tumour histology in postmenopausal women with large operable breast cancer, Br. J. Cancer 87 (2002) 334–338. W.R. Miller, J.M. Dixon, D.A. Cameron, T.J. Anderson, Biological and clinical effects of aromatase inhibitors in neoadjuvant therapy, J. Steroid Biochem. Mol. Biol. 79 (1–5) (2001) 103–107. W.R. Miller, S. White, J.M. Dixon, J. Murray, L. Renshaw, T.J. Anderson, Proliferation, steroid receptors and clinical/pathological response in breast cancer treated with letrozole, Br. J. Cancer 94 (2006) 1051–1056. M. Dowsett, I.E. Smith, S.R. Ebbs, J.M. Dixon, A. Skene, C. Griffith, I. Boeddinghaus, J. Salter, S. Detre, M. Hills, S. Ashley, S. Francis, G. Walsh, IMPACT Trialists: 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. 11 (2 Pt 2) (2005) 951s– 958s. Y. Tao, A. Klause, A. Vickers, K. Bae, M. Ellis, Clinical and biomarker endpoint analysis in neoadjuvant endocrine therapy trials, J. Steroid Biochem. Mol. Biol. 95 (1–5) (2005) 91–95. M. Dowsett, C. Harper-Wynne, I. Boeddinghaus, J. Salter, M. Hills, M. Dixon, S. Ebbs, G. Gui, N. Sacks, I. Smith, HER-2 amplification impedes the antiproliferative effects of hormone therapy in estrogen receptor-positive primary breast cancer, Cancer Res. 61 (23) (2001) 8452–8458. J.M. Dixon, J. Jackson, M. Hills, L. Renshaw, D.A. Cameron, T.J. Anderson, W.R. Miller, M. Dowsett, Anastrozole demonstrates clinical and biological effectiveness in oestrogen receptor-positive breast cancers, irrespective of the erbB2 status, Eur. J. Cancer 40 (18) (2004) 2742–2747. W.R. Miller, T.J. Anderson, S. White, D. Evans, A. Krause, J.M. Dixon, Growth factor signalling in clinical breast cancer and its impact on response to conventional therapies: the Edinburgh experience, Endocr. Relat. Cancer 12 (Suppl 1) (2005) S119–123. M.J. Ellis, Y. Tao, O. Young, S. White, A.D. Proia, J. Murray, L. Renshaw, D. Faratian, J. Thomas, M. Dowsett, A. Krause, D.B. Evans, W.R. Miller, J.M. Dixon, Estrogen-independent proliferation is present in estrogen-receptor HER2positive primary breast cancer after neoadjuvant letrozole, J. Clin. Oncol. 24 (19) (2006) 3019–3025. M.J. Ellis, A. Coop, B. Singh, L. Mauriac, A. Llombert-Cussac, F. Janicke, W.R. Miller, D.B. Evans, D. Dugan, C. Brady, E. Quebe-Fehling, M. Borgs, Letrozole is more effective neoadjuvant endocrine therapy than tamoxifen for ErbB1- and/or ErbB-2-positive, estrogen receptor-positive primary breast cancer: evidence from a phase III randomized trial, J. Clin. Oncol. 19 (No 18) (2001) 3808–3816. A Mackay, A. Urruticoechea, J.M. Dixon, T. Dexter, K. Fenwick, A. Ashworth, S. Drury, A. Larionov, O. Young, S. White, W.R. Miller, D.B. Evans, M. Dowsett, Molecular response to aromatase inhibitor treatment in primary breast cancer, Breast Cancer Res. 9 (3) (2007) R37. W.R. Miller, A. Larionov, L. Renshaw, T.J. Anderson, S. White, G. Hampton, J.R. Walker, S. Ho, A. Krause, D.B. Evans, J.M. Dixon, Aromatase inhibitors—gene discovery, J. Steroid Biochem. Mol. Biol. 106 (1–5) (2007) 130– 142. W.R. Miller, A.A. Larionov, L. Renshaw, T.J. Anderson, S. White, J. Murray, E. Murray, G. Hampton, J.R. Walker, S. Ho, A. Krause, D.B. Evans, J.M. Dixon, Changes in breast cancer transcriptional profiles after treatment with the aromatase inhibitor, letrozole, Pharmacogenet. Genomics 17 (10) (2007) 813– 826. W.R. Miller, A. Larionov, T.J. Anderson, J.R. Walker, A. Krause, D.B. Evans, J.M. Dixon, Predicting response and resistance to endocrine therapy: profiling patients on aromatase inhibitors, Cancer 112 (S3) (2007) 689–694. W.R. Miller, A. Larionov, L. Renshaw, T.J. Anderson, J.R. Walker, A. Krause, T. Singh, D.B. Evans, J.M. Dixon, Gene expression profiles differentiating between breast cancers clinically responsive or resistant to letrozole, J. Clin. Oncol. 27 (9) (2009) 1382–1387. W.R. Miller, Identification and mechanisms of endocrine resistance, Breast Cancer Res. 10 (Suppl 4) (2008) S19.
Contents lists available at ScienceDirect
Journal of Steroid Biochemistry and Molecular Biology journal homepage: www.elsevier.com/locate/jsbmb
Clinical, pathological, proliferative and molecular responses associated with neoadjuvant aromatase inhibitor treatment in breast cancer夽 W.R. Miller ∗ Breast Unit Research Group, University of Edinburgh, Western General Hospital, Edinburgh, United Kingdom
a r t i c l e
i n f o
Article history: Received 1 September 2009 Received in revised form 2 October 2009 Accepted 7 October 2009 Keywords: Aromatase inhibitors Response Breast cancer
a b s t r a c t Neoadjuvant treatment provides an exceptional setting in which to monitor clinical, pathological, proliferative and molecular responses to aromatase inhibitors. Sequential measurements of the primary tumour provide an accurate assessment of clinical changes and the relatively easy access to the tumour within the breast means that biopsies are available for histological and molecular measurements before and during treatment. Large randomised trials (P024 and IMPACT) together with informative nonrandomised studies have demonstrated clinical responses to third generation aromatase inhibitors in 40–70% of ER-positive tumours, rates generally significantly higher than observed with tamoxifen. Pathological responses in terms of reduced cellularity/increased fibrosis are also seen in 65–75% of cases. Whilst these are more often seen in clinically responding tumours, the correlation between clinical and pathological response is not absolute. A marked feature of treatment with third generation inhibitors is a reduction in cellular proliferation. Using Ki67 as a marker, this may be observed as early as 10–14 days into treatment. Reduction in proliferation with treatment may be seen in both clinically responding and non-responding tumours, although incidence and degree of effect are higher in responding cases. Aromatase inhibitor treatment frequently fails to reduce proliferation in tumours over-expressing HER-2. In terms of molecular events, aromatase inhibitor treatment is associated with changes in expression of genes classically associated with oestrogen regulation (KIAA0101, ZWINT, IRS1 and TFF1) and cell cycle progression, most notably mitotic phase proteins (CDC2, CCNB1 and CKS2). Changes occur both in clinically responding and non-responding tumours. Although expression of no individual gene correlates absolutely with response status, expression signatures can be produced which distinguish between responding and non-responding tumours. In terms of gene ontology, terms relating to macro-molecular biosynthesis, translation and structural components of ribosomes are significantly enriched. Finally, molecular signatures can be used to illustrate the relative homogeneity of responding tumours and the disparate nature of non-responding tumours suggesting multiple and diverse pathways associated with resistance. © 2009 Elsevier Ltd. All rights reserved.
1. Introduction Neoadjuvant treatment regimes in which aromatase inhibitors are given to patients with the primary tumour still within the breast provide excellent protocols by which to measure responses associated with therapy [1–3]. Firstly, clinical response may be accurately assessed by sequential measurement of the primary tumour. Additionally, because the cancer is invariably biop-
夽 Special Issue selected article from the IX International Aromatase Conference (Aromatase 2008) held at Shanghai, China, on October 13th–16th, 2008. ∗ Current address: 2 Stoneycroft Road, South Queensferry EH30 9HX, United Kingdom. E-mail address: [email protected]. 0960-0760/$ – see front matter © 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.jsbmb.2009.10.005
sied to confirm diagnosis and to determine oestrogen receptor (ER) status, tissue may be available to assess pathology, histology, proliferation and molecular phenotype/genotype and to relate these to clinical response. Because, in most protocols, patients come to definitive breast surgery after 3–4 months’ therapy, tumour is also available at this time and the effects of treatment may be measured on the same pathological and biological parameters. Experimental protocols have also been described in which additional biopsies have been taken after 10–14 days treatment so that early effects of therapy may be monitored and used to predict tumour response at 3–4 months [4,5]. The paper reviews data derived from the neoadjuvant use of aromatase inhibitors in terms of observations on clinical, pathological and proliferative responses. It also describes how these interrelate,
274
W.R. Miller / Journal of Steroid Biochemistry & Molecular Biology 118 (2010) 273–276
and may be used to understand the biological effects of aromatase inhibition in breast cancer and optimise patient management in those with the disease. 2. Clinical response Clinical response rates for third generation aromatase inhibitors varies between 40 and 75% [3]. Two major clinical trials (P024 and IMPACT) have randomised neoadjuvant use of aromatase inhibitors versus tamoxifen [6,7]. Benefits were greater for the aromatase inhibitors; in the P024 trial the difference was statistically significant (55% for letrozole versus 34% tamoxifen, p < 0.001) whereas that in IMPACT (anastrozole versus tamoxifen) did not reach statistical significance. 3. Pathological response Marked changes in tumour morphology and histology may be associated with aromatase inhibitor treatment. In oestrogen receptor-rich tumours effects on histological features include changes in cellularity, degree of fibrosis and histological grading. These changes in tumour morphology may be seen by 3–4 months as evidenced by a clear reduction in cellularity or increase in fibrosis [8,9]. Such pathological responses have been reported in 60–80% of tumours [10,11], although complete pathological responses are rarely seen within the 3–4 months timeframe [12]. Treatment may also reduce histological grade; mitotic score being the most commonly affected grading feature [11]. Although rates of clinical and pathological responses are similar and may be significantly associated, there is not exact concordance, and about a fifth of tumours display a discordant phenotype in which tumours either display a decrease in tumour volume without evidence of pathological changes or do not change in tumour size whilst displaying a decrease in cellularity [13]. 4. Proliferative response Marked decreases in tumour proliferation (as evidenced by reduced mitotic score/staining of Ki67) in ER+ve breast cancers have been observed following treatment with aromatase inhibitors [10,12]. These changes may be seen in about 80% of cases after 3 months’ treatment (with a median decrease of about 90%). As compared with neoadjuvant tamoxifen treatment the effects are generally more pronounced and seen in a greater proportion of cases [5,10]. However, reduced proliferation may be observed much earlier in treatment; for example, studies in which biopsies have been taken at 10–14 days of treatment show marked and significant decreases in the expression of Ki67 [13]. Although similar effects may be seen at these early time-points with tamoxifen, the effects are significantly greater with aromatase inhibitors [14]. Several studies have examined the patterns of changes in proliferation with time. The majority of tumours display decreases in proliferation at 10–14 days which are maintained or become more pronounced with treatment after 3 months. However, it is possible to observe in a minority of tumours other patterns in which (a) initial decrease in proliferation at 10–14 days is followed by recovery towards pre-treatment values at 3 months, (b) reduction in proliferation is seen at 3 months but not 10–14 days and (c) cases display little changes either at 10–14 days or 3 months [13]. Proliferative responses are positively correlated with both clinical and pathological response [13–15]. However, they are not equivalent and discordant phenotypes occur relatively frequently. For example, clinical response and cell cycle response to both tamoxifen and letrozole in the P024 trial were discordant in over one third of cases [15]. Similarly the degree of proliferative response
was not significantly different in cases clinically responding or resistant to anastrozole in the IMPACT trial [14]. However, it does appear that the relationship between proliferative response and pathological response is greater than that between proliferative response and clinical response [13], perhaps reflecting the closer association between two histological assessments over that between histology and tumour size. Interestingly, aromatase inhibitor treatment often fails to reduce proliferation in tumours over-expressing HER-2 [16–19]. 5. Molecular response Clear evidence of molecular responses has been elicited following treatment with aromatase inhibitors. For example, the progesterone receptor (PgR) which is classically regarded as a marker of oestrogenic activity is clearly reduced in 70–80% of breast cancers and, in about one half of cases, it may completely disappear [12,13]. Effects may be seen as early as 10–14 days and are consistent with the anti-oestrogenic effects of aromatase inhibitors. Loss of PgR expression may occur independently of pathological and clinical response [13,20]. Similar effects have been noted on the expression of other oestrogen regulated markers such as pS2 [20]. These effects may also be contrasted to those produced by neoadjuvant tamoxifen in which decreases in PgR and pS2 are much less frequent and, indeed, increases in expression may be often observed (reflecting the oestrogen agonist effects of this SERM). The use of microarray technology has great potential in identifying genes which are markers of molecular response and may be associated with/predict for clinical, pathological and proliferative response in breast cancers. There have been recent key publications [21–25]. For example, early changes in gene expression have been identified by comparing microarray analyses from paired tumour core biopsies taken before and after 14 days’ neoadjuvant treatment with letrozole [21–23]. These have confirmed that aromatase inhibitor treatment is associated with changes in expression of genes classically associated with oestrogen regulation such as KIAA0101, ZWINT, TFF1 and IRS1, and cell cycle progression most notably mitotic phase proteins such as CDC2, CCNB1 and CKS2. Although more frequent in clinically responding tumours, these changes also occur in clinically non-responsive tumours [24,25]. As a consequence, they are unlikely to be reliable predictors of clinical response. 6. Prediction of clinical response Because markers of oestrogen action and cell proliferation are likely to be imperfect predictors of clinical response, it is important to identify novel genes or expression profiles which have a greater utility. To do this, microarrays from pre- and 14-day treatment tumour biopsies have been compared in patients either responding or resistant to neoadjuvant letrozole treatment [22,24,25]. Expression of no single gene correlated absolutely with response status. However, expression signatures can be derived which distinguish between responding and non-responding tumours. This work has been published in a preliminary form [22,24] and is about to be published in full [25]. In brief, it has been possible to identify 205 covariables (69 baseline genes, 45 day 14 genes and 91 changes in gene expression between baseline and day 14) which are associated with clinical response status. In terms of molecular function, gene ontology analysis reveals ‘structural constituents of ribosomes’ is highly significantly enriched and, in terms of biological processes, terms such as macromolecule biosynthetic processes, cellular biosynthetic processes and translation are also significantly enriched. Complete linkage hierarchical clustering of cases on the basis of the covariables separates responders from non-responders
W.R. Miller / Journal of Steroid Biochemistry & Molecular Biology 118 (2010) 273–276
275
Fig. 1. Heatmap of the dissimilarity matrix between cases and corresponding dendrograms (top and left dendrograms are identical). Bar colours beneath the dendrogram code for response—blue for responding cases and purple non-responding. Colour tones in the dissimilarity matrix represent degree of similarity—red and orange indicate high similarity, yellow and white high dissimilarity. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of the article.)
into two distinct groups. It also illustrates that responding tumours comprise a relatively homogenous group clearly dissimilar genetically from the non-responding tumours (Fig. 1). 7. Markers of resistance Molecular studies may also be utilised to illustrate the relative heterogeneity of non-responding tumours [24,26]. For example, using classical markers of oestrogen regulation such as KAA0101, SERPINA3, IRS1, TFF3, TFF1 and proliferation such as CDC, CKS2, cyclin B1, TYMS, PCNA, it is possible to subdivide tumours resistant
to letrozole into three subgroups: (i) cases which are molecularly resistant both in terms of oestrogen markers and proliferation (a phenotype similar to the lack of gene expression changes seen in ER−ve tumours); (ii) oestrogen sensitive tumours which display a phenotype of a reduction in oestrogen regulated genes but no change or an increase in expression of cell cycle markers, and (iii) molecularly sensitive tumours displaying a reduction in both oestrogen and cell proliferation markers (see Fig. 2). These observations illustrate the disparate nature of non-responding tumours and suggest multiple and diverse pathways are associated with resistance.
Fig. 2. Changes in expression of oestrogen regulated and proliferation associated genes after 14 days adjuvant treatment with letrozole. Colour coding—green represents downregulation and red upregulation. Intensity of colour represents degree of change. Single column on left of figure represents changes in an ER−ve case; the other panels relate to 15 different clinically resistant ER+ve tumours—the left-hand panel illustrates a molecular resistant phenotype, the middle panel cases showing decreases in the expression of oestrogen regulated genes but not in cell cycle genes, and the right-hand panel molecular sensitivity in both oestrogen regulated and cell cycle associated gene expression. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of the article.)
276
W.R. Miller / Journal of Steroid Biochemistry & Molecular Biology 118 (2010) 273–276
8. Conclusions This review has illustrated how neoadjuvant treatment specifically using aromatase inhibitors presents a unique opportunity by which to assess the clinical, pathological, proliferative and molecular changes associated with this type of endocrine therapy in individual breast cancers. It is clear that whilst the different types of response are inter-related, there are relatively frequent discordant phenotypes where, for example, clinically responsive tumours fail to show pathological changes and clinically resistant tumours may show clear molecular responses. The understanding of the mechanisms behind these phenomena will undoubtedly help to rationalise treatment, to identify mechanisms of resistance and develop circumventing or synergistic treatment regimes; all of which should produce clinical benefits both in terms of higher response rates and more enduring tumour responses. Molecular phenotyping has also produced evidence of a relative homogeneity between tumours which are clinically responsive to aromatase inhibitors. Conversely, changes in expression of genes related to oestrogen signalling and proliferation can illustrate heterogeneity between resistant tumours. The implication of this heterogeneity is that multiple pathways and mechanisms are involved in resistance to aromatase inhibitors and that diverse strategies may be needed to circumvent such resistance. Genetic signatures are also being developed which can distinguish between responsive and resistant tumours. In future these may be used not only to predict cases in which to use aromatase inhibitors, but also to suggest alternative drug regimes which may be successful in resistant tumours. In summary, neoadjuvant therapy with aromatase inhibitors is not only producing clinical benefits in terms of reducing tumour size and allowing more conservative breast surgery, but it is providing fundamental knowledge on pathological, biological and molecular processes which will impact on the future use of these endocrine agents. References [1] M. Kaufmann, G.N. Hortobagyi, A. Goldhirsch, S. Scholl, A. Makris, P. Valagussa, J.U. Blohmer, W. Eiermann, R. Jackesz, W. Jonat, A. Lebeau, S. Loibl, W. Miller, S. Seeber, V. Semiglazov, R. Smith, R. Souchon, V. Stearns, M. Untch, G. von Minckwitz, Recommendations from an international expert panel on the use of neoadjuvant (primary) systemic treatment of operable breast cancer: an update, J. Clin. Oncol. 24 (12) (2006) 1940–1949. [2] J.M. Dixon, C.D. Love, C.O. Bellamy, D.A. Cameron, R.C. Leonard, H. Smith, W.R. Miller, Letrozole as primary medical therapy for locally advanced and large operable breast cancer, Breast Cancer Res. Treat. 66 (3) (2001) 191–199. [3] J.M. Dixon, J. Jackson, L. Renshaw, W.R. Miller, Neoadjuvant tamoxifen and aromatase inhibitors: comparisons and clinical outcomes, J. Steroid. Biochem. Mol. Biol. 86 (3–5) (2003) 295–299. [4] W.R. Miller, T.J. Anderson, D.B. Evans, A. Krause, G. Hampton, J.M. Dixon, An integrated view of aromatase and its inhibition, J. Steroid Biochem. Mol. Biol. 86 (2003) 413–421. [5] W.R. Miller, T.J. Anderson, S. White, A. Larionov, J. Murray, D. Evans, A. Krause, J.M. Dixon, Aromatase inhibitors: cellular and molecular effects, J. Steroid Biochem. Mol. Biol. 95 (2005) (2005) 83–89. [6] W. Eiermann, S. Paepke, J. Appfelstaedt, A. Llombart-Cussac, J. Eremin, J. Vinholes, L. Mauriac, M. Ellis, M. Lassus, H.A. Chaudri-Ross, M. Dugan, M. Borgs, Letrozole Neo-Adjuvant Breast Cancer Study Group, Preoperative treatment of postmenopausal breast cancer patients with letrozole: a randomized doubleblind multicenter study, Ann. Oncol. 12 (11) (2001) 1527–1532. [7] I.E. Smith, M. Dowsett, S.R. Ebbs, J.M. Dixon, A. Skene, J.U. Blohmer, S.E. Ashley, S. Francis, I. Boeddinghaus, G. Walsh, IMPACT Trialists Group, Neoadjuvant treatment of postmenopausal breast cancer with anastrozole, tamoxifen, or both
[8] [9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
in combination: the Immediate Preoperative Anastrozole, Tamoxifen, or Combined with Tamoxifen (IMPACT) multicenter double-blind randomized trial, J. Clin. Oncol. 23 (22) (2005) 5108–5116. J.M. Dixon, C.D.B. Love, L. Renshaw, et al., Lessons from the use of aromatase inhibitors in the neoadjuvant setting, Endocr. Relat. Cancer 6 (1999) 227–230. H. Sasano, T. Suzuki, T. Moriya, Pathology of breast cancer following neoadjuvant therapy, in: W.R. Miller, J.N. Ingle (Eds.), Endocrine Therapy in Breast Cancer, Marcel Dekker, New York, USA, 2002, pp. 213–222. W.R. Miller, J.M. Dixon, L. Macfarlane, et al., Pathological features of breast cancer response following neoadjuvant treatment with either letrozole or tamoxifen, Eur. J. Cancer 39 (4) (2002) 462–468. T.J. Anderson, J.M. Dixon, M. Stuart, et al., Effect of neoadjuvant treatment with anastrozole on tumour histology in postmenopausal women with large operable breast cancer, Br. J. Cancer 87 (2002) 334–338. W.R. Miller, J.M. Dixon, D.A. Cameron, T.J. Anderson, Biological and clinical effects of aromatase inhibitors in neoadjuvant therapy, J. Steroid Biochem. Mol. Biol. 79 (1–5) (2001) 103–107. W.R. Miller, S. White, J.M. Dixon, J. Murray, L. Renshaw, T.J. Anderson, Proliferation, steroid receptors and clinical/pathological response in breast cancer treated with letrozole, Br. J. Cancer 94 (2006) 1051–1056. M. Dowsett, I.E. Smith, S.R. Ebbs, J.M. Dixon, A. Skene, C. Griffith, I. Boeddinghaus, J. Salter, S. Detre, M. Hills, S. Ashley, S. Francis, G. Walsh, IMPACT Trialists: 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. 11 (2 Pt 2) (2005) 951s– 958s. Y. Tao, A. Klause, A. Vickers, K. Bae, M. Ellis, Clinical and biomarker endpoint analysis in neoadjuvant endocrine therapy trials, J. Steroid Biochem. Mol. Biol. 95 (1–5) (2005) 91–95. M. Dowsett, C. Harper-Wynne, I. Boeddinghaus, J. Salter, M. Hills, M. Dixon, S. Ebbs, G. Gui, N. Sacks, I. Smith, HER-2 amplification impedes the antiproliferative effects of hormone therapy in estrogen receptor-positive primary breast cancer, Cancer Res. 61 (23) (2001) 8452–8458. J.M. Dixon, J. Jackson, M. Hills, L. Renshaw, D.A. Cameron, T.J. Anderson, W.R. Miller, M. Dowsett, Anastrozole demonstrates clinical and biological effectiveness in oestrogen receptor-positive breast cancers, irrespective of the erbB2 status, Eur. J. Cancer 40 (18) (2004) 2742–2747. W.R. Miller, T.J. Anderson, S. White, D. Evans, A. Krause, J.M. Dixon, Growth factor signalling in clinical breast cancer and its impact on response to conventional therapies: the Edinburgh experience, Endocr. Relat. Cancer 12 (Suppl 1) (2005) S119–123. M.J. Ellis, Y. Tao, O. Young, S. White, A.D. Proia, J. Murray, L. Renshaw, D. Faratian, J. Thomas, M. Dowsett, A. Krause, D.B. Evans, W.R. Miller, J.M. Dixon, Estrogen-independent proliferation is present in estrogen-receptor HER2positive primary breast cancer after neoadjuvant letrozole, J. Clin. Oncol. 24 (19) (2006) 3019–3025. M.J. Ellis, A. Coop, B. Singh, L. Mauriac, A. Llombert-Cussac, F. Janicke, W.R. Miller, D.B. Evans, D. Dugan, C. Brady, E. Quebe-Fehling, M. Borgs, Letrozole is more effective neoadjuvant endocrine therapy than tamoxifen for ErbB1- and/or ErbB-2-positive, estrogen receptor-positive primary breast cancer: evidence from a phase III randomized trial, J. Clin. Oncol. 19 (No 18) (2001) 3808–3816. A Mackay, A. Urruticoechea, J.M. Dixon, T. Dexter, K. Fenwick, A. Ashworth, S. Drury, A. Larionov, O. Young, S. White, W.R. Miller, D.B. Evans, M. Dowsett, Molecular response to aromatase inhibitor treatment in primary breast cancer, Breast Cancer Res. 9 (3) (2007) R37. W.R. Miller, A. Larionov, L. Renshaw, T.J. Anderson, S. White, G. Hampton, J.R. Walker, S. Ho, A. Krause, D.B. Evans, J.M. Dixon, Aromatase inhibitors—gene discovery, J. Steroid Biochem. Mol. Biol. 106 (1–5) (2007) 130– 142. W.R. Miller, A.A. Larionov, L. Renshaw, T.J. Anderson, S. White, J. Murray, E. Murray, G. Hampton, J.R. Walker, S. Ho, A. Krause, D.B. Evans, J.M. Dixon, Changes in breast cancer transcriptional profiles after treatment with the aromatase inhibitor, letrozole, Pharmacogenet. Genomics 17 (10) (2007) 813– 826. W.R. Miller, A. Larionov, T.J. Anderson, J.R. Walker, A. Krause, D.B. Evans, J.M. Dixon, Predicting response and resistance to endocrine therapy: profiling patients on aromatase inhibitors, Cancer 112 (S3) (2007) 689–694. W.R. Miller, A. Larionov, L. Renshaw, T.J. Anderson, J.R. Walker, A. Krause, T. Singh, D.B. Evans, J.M. Dixon, Gene expression profiles differentiating between breast cancers clinically responsive or resistant to letrozole, J. Clin. Oncol. 27 (9) (2009) 1382–1387. W.R. Miller, Identification and mechanisms of endocrine resistance, Breast Cancer Res. 10 (Suppl 4) (2008) S19.