Adjuvant Exemestane Therapy After 5 Years of Tamoxifen: Rationale for the NSABP B-33 Trial

ABSTRACT: Tamoxifen (Nolvadex) has long been established as"standard" adjuvant therapy for receptor-positive, early-stage breastcancer. Results from clinical trials suggest that after approximately 5 years,tamoxifen maylose its effectiveness and may even become harmful if not stopped. At the timeof tamoxifen discontinuation, "seemingly" disease-free patients maystill have residual micrometastatic tumor cells. In a proportion of suchpatients, these cells may still be responsive to tamoxifen and thus could growas a result of stopping the drug. The majority of clinical information, however,suggests that by 5 years of therapy, a greater proportion of patients will havemicrometastatic tumor cells that have become resistant to tamoxifen and will bestimulated by continuation of the drug for a longer time. Because in both suchcases the micrometastases are hormonally sensitive, a reasonable approach—inaddition to stopping tamoxifen—is to decrease the level of estrogenicstimulation by introducing an aromatase inhibitor. The National SurgicalAdjuvant Breast and Bowel Project is launching a clinical trial (NSABP B-33) toevaluate the sequential administration of 2 years of exemestane (Aromasin), asteroidal aromatase inactivator, after 5 years of tamoxifen in postmenopausal,receptor-positive patients who are disease-free at the time of tamoxifendiscontinuation. [ONCOLOGY 15(Suppl 7):35-39, 2001]

Introduction

Results from several clinical trials have clearly demonstratedsubstantial benefit from tamoxifen (Nolvadex) administration in stage I and IIhormonally responsive breast cancer.[1-9] None of the newly developedantiestrogens has been shown to be superior to tamoxifen in the adjuvantsetting. Thus, tamoxifen remains the "standard" adjuvant hormonaltherapy for these patients.

Although the optimal duration of tamoxifen administrationcontinues to be debated,[10] the majority of researchers agree that—at leastin some patients—there is a time interval beyond which tamoxifen loses itseffectiveness and may even become harmful if continued.

Evidence supporting this hypothesis is provided by studiesevaluating various intervals of tamoxifen administration in the adjuvantsetting; it is also corroborated by observations in the advanced-diseasesetting, where some patients progress after responding to tamoxifen therapy onlyto demonstrate a "withdrawal response" following tamoxifendiscontinuation.[11]

Optimal Duration of Adjuvant Tamoxifen Therapy

Short Duration Therapy: < 5 Years

Initial trials of adjuvant tamoxifen therapy[2-5] comparedtamoxifen given for 1 to 2 years to no adjuvant therapy. Results showedreductions in recurrence rates, mostly during the period of treatment, withlittle or no additional benefit afterward. These data, along with preclinicalinformation showing that tamoxifen was cytostatic rather than cytocidal,supported that the benefit of tamoxifen might be augmented by a longer durationof administration.

A Swedish multicenter trial reported in the early 1980s[7]compared 2 years with 5 years of adjuvant tamoxifen in postmenopausal women withearly-stage, node-negative or node-positive breast cancer. Of 3,887 patientsentered in the trial, 91% remained alive and recurrence-free at 2 years and wereincluded in the 2-year vs 5-year comparison. Ten years after initialrandomization, disease-free and overall survival were significantly improvedamong patients receiving the longer tamoxifen administration (disease-freesurvival: 73% vs 67%, P = .009; overall survival: 80% vs 74%, P = .03).

Similar results were obtained from a British trial,[8] in which2,937 patients who were recurrence-free after receiving 2 years of tamoxifenwere randomly assigned to no further therapy or 3 additional years of treatment.With a median follow-up of 2 years, a statistically significant benefit in timeto disease relapse was seen in the patients receiving the longer treatmentduration (relative risk: 0.81, 95% confidence interval [CI] = 0.69%-0.98%).Fewer deaths were also observed in the group receiving the longer treatment, butthe difference was not statistically significant (relative risk: 0.89, 95% CI =0.69%-1.15%).

Other trial results also support a longer duration (up to 5years) of tamoxifen administration, including the Early Breast Cancer Trialists’Collaborative Group Meta-Analysis[12] and a nonrandomized trial by NSABP (theNational Surgical Adjuvant Breast and Bowl Project),[9] which compared 3-year vs2-year therapy with adjuvant tamoxifen.

Long Duration Therapy: > 5 Years

The next obvious question was whether a longer duration oftamoxifen administration (ie, more than 5 years) would result in furtherimprovement in disease-free and overall survival. To address this question, theNSABP rerandomized patients who had participated in a trial comparing adjuvanttreatment with tamoxifen vs placebo and who were recurrence-free after 5 yearsof tamoxifen, to 5 more years of tamoxifen or 5 years of placebo.[13] Through 4years after rerandomization, patients receiving placebo had better disease-freesurvival (92% vs 86%, P = .003) and overall survival (96% vs 94%, P = .08) thanthose receiving tamoxifen.[13] Thus, this trial demonstrated that node-negativepatients derived no additional benefit from continuing tamoxifen therapy beyond5 years and that continuing tamoxifen past that interval was possiblydetrimental.

Comparable results were obtained in a Scottish trial[14] ofsimilar design in terms of continuing vs stopping tamoxifen therapy after 5years. With 342 patients reassigned, and after a median follow-up of 6 years, anonsignificant trend of a higher recurrence rate was noted for patients whocontinued tamoxifen (relative risk, 1.27; 95% CI = 0.87%-1.85%).

In contrast, a smaller Eastern Cooperative Oncology Group (ECOG)study compared tamoxifen administration for 5 years vs more than 5 years in 193node-positive patients. Results demonstrated a nonsignificant trend towardsfewer recurrences in patients receiving longer tamoxifen administration (15 vs23).[15]

Mechanisms of Tamoxifen Resistance

The exact mechanism of tamoxifen resistance remains elusive.Experimental and clinical evidence suggests that more than one mechanism may beinvolved, including the development of partial agonistic activity of tamoxifenwith stimulation of tumor regrowth,[11,13,14,16,17] alteration or mutation ofthe estrogen-receptor,[16,18-20,21] clonal selection of an estrogen-receptor-negativephenotype,[16,20,22] development of metabolic tolerance resulting in inadequateintra-tumoral tamoxifen concentrations,[16,20,22] and increased gene expressionand growth factor production by tumor cells possibly leading to autocrinestimulation.[23,24]

Recently, development of estrogen hypersensitivity has beenpostulatedas another mechanism of tamoxifen resistance.[25] According to this hypothesis,deprivation of estrogen receptors from estrogen, induced by the administrationof tamoxifen, can cause enhanced estrogen sensitivity. Thus, further reductionof estrogen levels by an aromatase inhibitor might be effective by furtherdepriving hypersensitive tumor cells of estrogen levels necessary for theirgrowth.

In patients who have disease recurrence while taking adjuvanttamoxifen therapy, several of these mechanisms may contribute to development oftamoxifen resistance. The preponderant mechanism(s) will dictate whetheradditional hormonal manipulations might provide further benefit. In theadvanced-disease setting, patients who develop acquired resistance after aninitial response have an approximate 50% chance of responding to anotherhormonal agent.[22]

Sequential Administration of an Aromatase Inhibitor

Tamoxifen-Resistant Cells

Based on the above information, adjuvant tamoxifen is given forapproximately 5 years to most patients— particularly those with negativenodes. Thus, the majority of patients are disease-free at the time of tamoxifendiscontinuation. Some of these seemingly disease-free patients, however, harborresidual or micrometastatic tumor cells even after several years of tamoxifentherapy. In a proportion of these patients the residual or micrometastatic tumorcells may still be responsive to tamoxifen and may be allowed to grow andproliferate by tamoxifen discontinuation.

The clinical data, however, support the notion that in a greaterproportion of patients the residual or micrometastatic tumor cells mayprogressively become resistant to tamoxifen and could even be stimulated by thedrug if it were to be continued for a longer time. Thus, continuing tamoxifenmay benefit those patients with tamoxifen-responsive tumor cells but would bedetrimental to the group whose cells have become resistant to the agent. Sincethe residual or micrometastatic tumor cells in both of these groups arehormonally responsive, reducing the level of estrogenic stimulation at the timeof tamoxifen discontinuation is a logical approach to further reduce theprobability of recurrence (Figure 1).

The Role of Aromatase Inhibitors

Aromatase inhibitors act systemically to prevent estrogenbiosynthesis by inhibiting the enzyme aromatase, which catalyses the conversionof adrenal and ovarian androgen to estrogen. These compounds have been shown tobe effective in postmenopausal women with hormone-responsive metastatic breastcancer.

Aminoglutethimide (Cytadren), a first-generation aromataseinhibitor, was effective in the adjuvant breast cancer setting,[26] but waspoorly tolerated and was supplanted by the well-tolerated second-generationaromatase inhibitor 4-OH-androstenedione (4-OHA, formestane). However, thiscompound suppressed plasma estradiol levels to only one-third of baseline levelsand required parenteral administration.[27,28] Several years later,third-generation aromatase inhibitors were developed that fell into twoprincipal categories: (a) nonsteroidal aromatase inhibitors, such as fadrozole,vorozole (Rivizor), letrozole (Femara), and anastrozole (Arimidex), and (b)steroidal aromatase inhibitors, such as exemestane (Aromasin).

No information currently exists on whether additional hormonalinterventions will be beneficial in breast cancer patients who arerecurrence-free after discontinuation of tamoxifen therapy. Ample information,however, indicates that sequential administration of aromatase inhibitors topatients who recur during or after tamoxifen therapy results in significantantitumor activity.[27-30] The downside of waiting and treating when recurrenceoccurs is that by then, an additional proportion of tumors may have becomehormone-resistant and another proportion would remain hormone-sensitive only fora short time. Thus, there are theoretical advantages in attempting to reduce therisk of subsequent recurrence in patients who remain disease-free aftercompletion of adjuvant tamoxifen therapy by initiating therapy with an aromataseinhibitor.

Exemestane in Adjuvant Setting

Exemestane is a very potent, orally active, selective andirreversible, steroidal aromatase inactivator. In contrast to the nonsteroidalinhibitors of aromatase that bind reversibly to the heme part of the enzyme,steroidal inactivators mimic the natural substrate androstenedione and bindcovalently (irreversibly) to the substrate-binding site of aromatase. As aresult of the inactivation, the amount of aromatase present is markedlydecreased.[31]

Phase I-II Studies

Exemestane suppresses plasma estrogen levels to 6% to 11% ofpretreatment levels,[32,33] thus demonstrating activity comparable to thatobserved with third-generation nonsteroidal aromatase inhibitors such asanastrozole or letrozole, and significantly more pronounced than that offormestane. In phase I studies, a total of 137 patients were treated withexemestane at various daily doses (up to 600 mg). No maximum tolerated dose wasdetermined, while 25 objective responses were observed (18%).[34,35]

Exemestane was evaluated in five phase II trials, in whichapproximately 700 postmenopausal women with metastatic breast cancer receivedoral exemestane at 25 mg daily.[36-40] Patients in two of the studies hadexperienced failure to tamoxifen therapy; patients in two other studies hadexperienced failure to tamoxifen and megestrol acetate (Megace); and patients inanother study had failed tamoxifen and an aromatase inhibitor.

Response rates were 28% to 31% for patients experiencing failureafter tamoxifen and 7% to 13% for those experiencing failure after tamoxifen andanother hormonal therapy. Rates of overall success ranged between 47% and 59% inpatients experiencing tamoxifen failure and 24% to 30% in those experiencingfailure of two hormonal therapies. Time to progression ranged from 24 to 25weeks and 9 to 16 weeks, respectively. These phase II studies have shown thatexemestane is an active and well-tolerated hormonal agent in postmenopausalpatients with metastatic breast cancer.

Efficacy and Safety

A phase III, double-blind, randomized, multicenter studyevaluated the efficacy and safety of exemestane vs megestrol acetate inpostmenopausal women with progressive advanced breast cancer who had experiencedfailure of tamoxifen therapy.[41] A total of 769 patients were randomly assignedto receive oral exemestane25 mg daily or oral megestrol acetate 40 mg qid.

There were no statistically significant differences in overallobjective response rates between exemestane vs megestrol acetate (15% vs 12.4%,respectively) or overall success rates (37.4% vs 34.6%, respectively). However,patients receiving exemestane had a significantly longer median survival time(median not reached) than those receiving megestrol acetate (123.4 weeks; P =.039). They also had significantly better median duration of overall success(60.1 vs 49.1 weeks; P = .025), time to tumor progression (20.3 vs 16.6 weeks; P= .037), and time to treatment failure (16.3 vs 15.7 weeks; P = .042). Toxicityinformation demonstrated that exemestane was well tolerated. The overallincidence of adverse events was 39.1% in patients receiving exemestane and 45.8%in those receiving megestrol acetate.

Possible Long-Term Adverse Effects

A concern regarding sequential therapy with aromatase inhibitorsin breast cancer patients who are disease-free following completion of 5 yearsof tamoxifen therapy is the possible long-term adverse effects of aromataseinhibitors on bone mineral density and blood lipids and lipoproteins. This isimportant because the majority of patients who are recurrence-free followingtamoxifen therapy will remain so without additional treatment. Therefore, inselecting an aromatase inhibitor for use in the adjuvant setting, the effects onbone mineral density and blood lipids and lipoproteins are of particularsignificance. The weak androgenic effect of exemestane may provide a theoreticaladvantage in regard to those parameters compared with the class of nonsteroidalaromatase inhibitors such as anastrozole and letrozole, but this remains to beseen in adjuvant therapy trials.

The NSABP B-33 Trial

The NSABP is about to initiate protocol B-33, a randomized,placebo-controlled, double-blind trial evaluating the effects of exemestane instage I and II, postmenopausal, estrogen-receptor-positive and/orprogesterone-receptor-positive breast cancer patients who have completed 5years of tamoxifen therapy (Figure 2).

Trial Goals

The primary aim of the trial is to determine whether theadministration of exemestane for 2 years after 5 years of tamoxifen therapy willprolong disease-free survival compared with a placebo. Secondary aims are todetermine whether the administration of exemestane will prolong overall survivaland time to treatment failure. Tertiary aims are to evaluate the effect oftamoxifen withdrawal on bone mineral density and to determine if exemestane hasany additional effect on the rate of bone loss resulting from tamoxifendiscontinuation.

Eligibility

Eligible patients must have completed approximately 5 years ofadjuvant tamoxifen therapy and must be postmenopausal and disease-free at thetime of randomization. They must have a life expectancy of at least 10 years,excluding their cancer diagnosis. The interval between completion of tamoxifentherapy and randomization must be no more than 6 months. Eligible patients musthave previously undergone either a modified radical mastectomy or a lumpectomyand axillary dissection.

At the time of their breast cancer diagnosis, patients may havereceived adjuvant or neoadjuvant chemotherapy either concomitantly with orbefore initiation of tamoxifen therapy. They may have received postmastectomylocoregional radiotherapy or postlumpectomy regional radiotherapy. Patients whounderwent lumpectomy must have received breast radiotherapy postlumpectomy.

In addition, eligible patients must have determinations of bothestrogen and progesterone receptors performed on the primary tumor at the timeof the original resection. The primary tumor must be estrogen-receptor positiveand/or progesterone-receptor positive.

Evaluation

The effects of exemestane and tamoxifen withdrawal on bone willbe evaluated in the entire study population by collecting data on symptomaticfractures, height, and total serum alkaline phosphatase. In addition, a patientsubgroup will be prospectively evaluated with serial measurements of both bonemineral density and bone biochemical markers. Bone mineral density of the lumbarspine and hip and total body calcium will be measured by dual-energy x-rayabsorptiometry at baseline, 1 year, 2 years, and 3 years after treatmentinitiation. Blood samples for measurement of biochemical markers of boneturnover will be collected at baseline, 6 months, 1 year, 2 years, and 3 yearsafter treatment initiation.

The study will also evaluate the effect of exemestane on bloodlipids by obtaining a fasting lipid panel consisting of total cholesterol, HDLcholesterol, and triglyceride levels before, during, and after completion of thestudy therapy in all patients registered on the bone mineral density substudy.

Assuming that half the patients accrued to this trial will benode-negative and half will be node-positive, a total of 3,000 patients will beneeded in order to have 80% power to detect a 21% reduction in hazard rate usinga two-sided 0.05-level log-rank test.

Summary and Outlook

Although a majority of patients are disease-free when theydiscontinue tamoxifen after approximately 5 years, some of these patients harborresidual or micrometastatic tumor cells. Ample information indicates thataromatase inhibitors sequentially administered to patients who recur during orafter tamoxifen therapy results in significant antitumor activity. Thisinformation has lent support to the theory that initiating therapy witharomatase inhibitors in patients without evidence of disease after completingadjuvant therapy can reduce the risk of subsequent recurrence.

Exemestane, a potent aromatase inactivator, rather than aninhibitor, will soon be tested in stage I and II, postmenopausal,estrogen-receptor-positive and/or progesterone-receptor- positive patientswho have completed 5 years of tamoxifen therapy. The major aims of the NSABPrandomized, placebo-controlled study are to determine whether administeringexemestane for 2 years, as compared to a placebo, will prolong disease-free andoverall survival and time to treatment failure.

References:

1. Early Breast Cancer Trialists’ Collaborative Group:Tamoxifen for early breast cancer: An overview of the randomized trials. Lancet351:1451-1467, 1998.

2. Baum M, Brinkley DM, Dossett JA, et al: Controlled trial oftamoxifen as a single adjuvant agent in the management of early breast cancer:Analysis at 8 years by Nolvadex Adjuvant Trial Organization. Br J Cancer57:608-611, 1988.

3. Bartlett K, Eremin O, Hutcheon A, et al: Adjuvant tamoxifenin the management of operable breast cancer. The Scottish Trial. Lancet2:171-175, 1987.

4. Ribeiro G, Swindell R: The Christie Hospital adjuvanttamoxifen trial—Status at 10 years. Br J Cancer 57:601-603, 1988.

5. Cancer Research Campaign Breast Cancer Trials Group: Theeffect of adjuvant tamoxifen: The latest results from the Cancer ResearchCampaign Adjuvant Breast Trial. Eur J Cancer 28A(4/5):904-907, 1992.

6. Fisher B, Constantino J, Redmond C, et al: A randomizedclinical trial evaluating tamoxifen in the treatment of patients withnode-negative breast cancer who have estrogen-receptor-positive tumors. N EnglJ Med 320:479-484, 1989.

7. Swedish Breast Cancer Cooperative Group: Randomized trial of2 versus 5 years of adjuvant tamoxifen in postmenopausal early stage breastcancer. J Natl Cancer Inst 88:1543-1549, 1996.

8. Current Trials Working Party on the Cancer Research CampaignBreast Trials Group: Preliminary results from the Cancer Research Campaign trialevaluating tamoxifen duration in women age 50 years or older with breast cancer.J Natl Cancer Inst 88:1834-1839, 1996.

9. Fisher B, Brown A, Wolmark N, et al: Prolonging tamoxifentherapy for primary breast cancer. Findings from the National Surgical AdjuvantBreast and Bowel project clinical trial. Ann Intern Med 106:649-654, 1987.

10. Peto R: Five years of tamoxifen—or more? J Natl CancerInst 88:1791-1793, 1996.

11. Howell A, Dodwell DJ, Anderson H, et al: Response afterwithdrawal of tamoxifen and progestogens in advanced breast cancer. Ann Oncol3:611-617, 1992.

12. Early Breast Cancer Trialists’ Collaborative Group:Systemic treatment of early breast cancer by hormonal, cytotoxic, or immunetherapy—133 randomized trials involving 31,000 recurrences and 24,000 deathsamong 75,000 women. Lancet 339:71-85, 1992.

13. Fisher B, Dignam J, Bryant J, et al: Five versus more than 5years of tamoxifen therapy for breast cancer patients with negative lymph nodesand estrogen-receptor-positive tumors. J Natl Cancer Inst 88:1529-1542, 1996.

14. Stewart HJ, Forrest AP, Everington D, et al: Randomizedcomparison of 5 years of adjuvant tamoxifen with continuous therapy for operablebreast cancer. Br J Cancer 74:297-299, 1996.

15. Tormey DC, Gray R, Falkson HC (for the Eastern CooperativeOncology Group): Postchemotherapy adjuvant tamoxifen therapy beyond 5 years inpatients with lymph node positive breast cancer. J Natl Cancer Inst88:1828-1833, 1996.

16. Johnston SR: Acquired tamoxifen resistance in human breastcancer—Potential mechanisms and clinical implications. Anticancer Drugs8:911-930, 1997.

17. De Friend DJ, Anderson E, Bell J, et al: Effects of4-hydroxytamoxifen and a pure novel antiestrogen (ICI 182780) on the clonogenicgrowth of human breast cells in vitro. Br J Cancer 70:204-211, 1994.

18. Tonetti DA, Jordan VC: The role of estrogen-receptormutations in tamoxifen-stimulated breast cancer. J Steroid Biochem Mol Biol62:119-128, 1997.

19. Madsen MW, Reiter BE, Larsen SS, et al: Estrogen-receptormessenger RNA splice variants are not involved in antiestrogen resistance insublimes of MCF-7 human breast cancer cells. Cancer Res 57:585-589, 1997.

20. Lykkesfeldt AE: Mechanisms of tamoxifen resistance in thetreatment of advanced breast cancer. Acta Oncol 35(suppl 5):9-14, 1996.

21. Zhang QX, Borg A, Wolf DM, et al: An estrogen-receptormutant with strong hormone-independent activity from a metastatic breast cancer.Cancer Res 57:1244-1249, 1997.

22. Dowsett M: Endocrine resistance in advanced breast cancer.Acta Oncol 35(suppl 5):91-95, 1996.

23. Arteaga CL, Koli KM, Dugger TC, et al: Reversal of tamoxifenresistance of human breast carcinomas in vivo by neutralizing antibodies totransforming growth factor-beta. J Natl Cancer Inst 91:46-53, 1999.

24. McLeskey SW, Zhang L, El-Ashry D, et al: Tamoxifen-resistantfibroblast growth factor-transected MCF-7 cells are cross resistant in vivo toantiestrogen ICI 182,780 and two aromatase inhibitors. Clin Cancer Res4:697-711, 1998.

25. Yue W, Santen R: Aromatase inhibitors: Rationale for usefollowing antiestrogen therapy. Semin Oncol 23(suppl 9):21-27, 1996.

26. Coombes RC, Powles TJ, Easton D, et al: Adjuvantaminoglutethimide therapy for postmenopausal patients with primary breastcancer. Cancer Res 47:2494-2497, 1987.

27. Coombes RC, Goss P, Dowsett M, et al:4-Hydroxyandrostenedione in treatment of postmenopausal patients with advancedbreast cancer. Lancet 2:1237-1239, 1984.

28. Goss PE, Powles TJ, Dowsett M, et al: Treatment of advancedpostmenopausal breast cancer with aromatase inhibitor 4-hydroxyandrostenedione—PhaseII report. Cancer Res 46:4823-4826, 1986.

29. Ibrahim NK, Buzdar AU: Aromatase inhibitors: Current status.Am J Clin Oncol 18(5):407-417, 1995.

30. Miller WR: Aromatase inhibitors. Endocrine-Related Cancer3:65-79, 1996.

31. Brueggemeier RW: Aromatase inhibitors—Mechanisms ofsteroidal inhibitors. Breast Cancer Res Treat 30:31-42, 1994.

32. Di Salle E, Ornati G, Paridaens R, et al: Preclinical andclinical pharmacology of the aromatase inhibitor exemestane (FCE 24304), inMotta M, Serio M (eds): Sex Hormones and Antihormones in Endocrine DependentPathology: Basic and Clinical Aspects, pp 303-309. New York, Elsevier Science,1994.

33. Di Salle E, Briatico G, Giudice D, et al: Novel aromataseand 5 alpha-reductase inhibitors. J Steroid Biochem Mol Biol 49:289-294, 1994.

34. Evans TR, Di Salle E, Ornati G, et al: Phase I and endocrinestudy of exemestane (FCE 24304), a new aromatase inhibitor in postmenopausalwomen. Cancer Res 52:5933-5939, 1992.

35. Lemnens J, Paridaens R, Thomas J, et al: Phase I study ofexemestane (FCE 24304), a new aromatase inhibitor in metastatic breast cancer.Ann Oncol 5(suppl 5):134, 1994.

36. Jones S, Belt R, Cooper B, et al: A phase II study ofantitumor efficacy and safety of exemestane as second-line hormonal treatment ofpostmenopausal patients with metastatic breast cancer refractory to tamoxifen(abstract 436). Breast Cancer Res Treat 50:304, 1998.

37. Kvinnsland S, Anker G, Dirix LY, et al: Activity ofexemestane, an irreversible, oral, aromatase inhibitor in metastaticpostmenopausal breast cancer patients failing tamoxifen (abstract 408). Eur JCancer 34(suppl 5):S91, 1998.

38. Jones S, Vogel C, Arkhipov A, et al: Multicenter, phase IItrial of exemestane as third-line hormonal therapy of postmenopausal women withmetastatic breast cancer. J Clin Oncol 17:3418-3425, 1999.

39. Jones S, Chang A, Lusch C, et al: A phase II confirmatorystudy of antitumor efficacy and safety of exemestane as third-line hormonaltreatment of postmenopausal patients with metastatic breast cancer refractory totamoxifen and Megace (abstract 437). Breast Cancer Res Treat 50:305, 1998.

40. Lonning PE, Bajetta E, Murray R, et al: A phase II study ofexemestane in metastatic breast cancer patients failing non-steroidal aromataseinhibitors (abstract 435). Breast Cancer Res Treat 50:304, 1998.

41. Kaufmann M, Bajetta E, Dirix LY, et al: Exemestane issuperior to megestrol acetate after tamoxifen failure in postmenopausal womenwith advanced breast cancer: Results of a phase III randomized double-blindtrial. J Clin Oncol 18:1399-1411, 2000.