Breast Cancer

 

The

Oncologist

®

ASCO 2001: Critical Commentaries

Breast Cancer

  y    h   p   r   u    M  .    P   n   a   e    S    ©

IRENE KUTER Massachusetts General Hospital, Boston, Massachusetts, USA

Key Words. Sentinel node micrometastases · Radiation after lumpectomy · Amenorrhea · Tamoxifen and BRCA mutations · HER-2/FISH · Tamoxifen resistance

A BSTRACT Several interesting aspects of breast cancer were covered at this year’s American Society of Clinical Oncology meeting. Sentinel lymph node (SN) mapping is now in widespread use, in concert with the general trend toward trying to decrease the morbidity of breast cancer surgery. With every advance, however, comes new challenges, and there was a timely presentation from Giuliano’s group addressing the controversial issue of how to interpret the presence of cells in the SN seen only with keratin stains but not by routine hematoxylin and eosin stains. Two abstracts addressed the issue of whether for certain women with invasive breast cancer radiation therapy could be omitted after lumpectomy. Another interesting topic related to hormonal issues in the adjuvant treatment of premenopausal women. An analysis from the ZIPP-TRIAL reported on bone marrow density studies in young women given two years of ovarian suppression in the adjuvant setting: it seems that the loss of bone density may be reversible and, more interestingly, may be prevented with

the prognostic significance of drug-induced amenorrhea in young women treated with adjuvant chemotherapy and at the efficacy of ovarian suppression during chemotherapy in preserving fertility. In an unpublicized presentation,  Mary-Claire King presented very interesting results from the National Surgical Adjuvant Breast and Bowel Project Breast Cancer Prevention Trial suggesting that tamoxifen may be an effective chemopreventive drug for women with BRCA2, but not BRCA1, mutations. Two important presentations re-analyzed the outcome of the pivotal trials using Herceptin to treat metastatic breast cancer and nicely show that FISH analysis of HER-2 overexpression is a more accurate indicator of response to Herceptin than immunohistochemical staining. Finally, there were two interesting presentations related to tamoxifen resistance which may be relevant clinically, pertaining to subsequent raloxifene use and the interaction of the estrogen receptor and EGF receptor pathways, respectively. The Oncologist 

concurrent tamoxifen. Two other presentations looked at

2001;6:338-346 

MANA ANAGEM GEMENT ENT OF EARLY STAGE BREAST CANCER

sentinel lymph nodes (SNs) detected only by immunohistochemistry (IHC). Between January 1992 and April 1999, 696 patients underwent SN mapping and were classified into four groups: A) SN negative by both hematoxylin and eosin (H&E) staining (n = 425); B) SN H&E –/IHC+ (n = 56); C) SN H&E+, micrometastases ≤2 mm (n = 76), and D) SN H&E+, macrometastases >2 mm ( n = 139). With a median follow-up of 38 months, the size of SN metastases was a significant predictor of disease-free survival (DFS) but not

Clinical Significance of Axillary Micrometastases in Breast Cancer: How Small is Too Small?  NM

Hansen,  BJ Grube, W Te, ML Brennan, R Turner, AE Giuliano (A BSTRACT  91)  91). This prospective study was designed to determine the significance in terms of survival of micrometastases in

Correspondence: Irene Kuter, M.D., D. Phil., Massachusetts General Hospital, Cox 640, Hematology-Oncology Unit, 100 Blossom Street, Boston, 02114, USA. Telephone: 617-726-8743; Fax: e-mail: [email protected]  Rec  Receiv eived ed July 13, July 1Massachusetts 3, 2001 2001;; accep accepted ted for pub publica lication tion Jul Julyy 23, 23 , 2001. 2001. ©Alp ©AlphaM haMed ed617-724-3166; Presss 1083Pres 1083-715 7159/2 9/2001 001/$5 /$5.00 .00/0. /0.

The Oncologist 2001;6:338-346   www.TheOncologist.com

 

 

339

Breast Cancer

Table 1. Sentinel lymph node status and clinical outcome Group

H&E

IHC

5-Year DFS (95% CI)

5-Year OS (95% CI)

I

–

–

95.1% (92.6, 97.5)

99.7% (99.1, 100)

II

–

+

98.3% (95.0, 100)

100%

III

+ micromets

94.5% (86.9, 100)

100%

IV

+ macromets

75.2% (65.2, 85.3)

96.5% (92.4, 100)

CI = confidence interval

overall survival (OS). However, as seen in Table 1, there was no significant difference in DFS or OS between SN – patients and those with SNs that were IHC+ but H&E –. The authors conclude that IHC should not be performed routinely on the SN nor should treatment decisions be made until results of large multicenter studies such as the American College of Surgeons Oncology Group (ACSOG) Z0010 are reported. Commentary This study addresses one of the issues most frequently discussed in breast conferences around the country: do a few cells in an SN detected only by IHC affect prognosis sufficiently to influence decision-making regarding adjuvant chemotherapy? It is well known that the status of the lymph nodes is the most important prognostic factor in breast cancer, and there is a continuum of worsening prognosis associated with each incrementally involved lymph node. It would seem intuitive, therefore, that finding metastatic cells by any technique would connote a worse prognosis than would be found in patients with totally negative nodes. Why then the controversy? In a commentary on this abstract,  Dr.  Laura Esserman pointed out that even in ductal carcinoma in situ, 7% to 13% of patients have been found to have pos-

itive SNs, yet in this disease the mortality is only approximately 1%. This illustrates the possibility that, in certain patients, the invasive tumors may have shed cells to the SNs but that these cells do not have the capacity to proliferate and hence will not adversely affect survival.  Dr. Esserman cited research regarding whether micrometastases should upstage a patient [1]. The trials were equally divided for and against the issue. So where do we stand now with respect to metastatic cells in the SN detected only by IHC? Although this trial so far shows no adverse effects on survival, it is certainly premature to draw definitive conclusions from such small numbers with short follow-up. Any adverse impact on survival is likely to be very small. We need to await the results of studies such as the ACSOG Z0010. In this large, multicenter trial, patients undergoing SN mapping are having the keratin stains done at a central institution. Treating surgeons and medical oncologists will be blinded to the

results. Only then can a definitive assessment of the true significance of IHC+ cells be made. Preliminary Results of a Randomized Study of  Tamoxifen ± Breast Radiation Radiation in T1/2 T1/2 N0 Disease in Women Over 50 Years of Age.  A Fyles,

D McCready, L  Manchul, M Trudeau, I Olivotto, P Merante, M Pintilie,  L Weir (A BSTRACT  92)  92). Comparison of Lumpectomy Plus Tamoxifen With and Without Radiotherapy (RT) in Women 70 Years of  Age or Older Who Have Clinical Stage I, Estrogen Receptor Positive (ER+) Breast Carcinoma. KS Hughes,

 L Schnaper, D Berry, C Cirrincione, B McCormick, B Shank, JD Lu, T Smith, B Smith, C Shapiro, W Wood, C  Henderson, L Norton (A BSTRACT  93).  93). In these two studies, the question being addressed is whether there is a subset of women who could forgo radiation therapy (RT) after excision of an invasive cancer. In both studies, tamoxifen was given to all the women, but only half  received breast radiation after lumpectomy. In the Fyles study, 638 women over 50 years old with T1 or T2, N0 breast cancers participated. Ipsilateral breast tumor relapse (IBTR) was seen in 22 of 384 (5.7%) patients in the tamoxifen arm compared with 2 of 385 (0.5%) in the tamoxifen plus radiation arm (Table 2). There was no difference in the relapse rate at regional or distant sites or in the contralateral breast, and survival was not influenced by RT. The authors concluded that tamoxifen and radiation therapy significantly lowered the rate of breast relapse compared with tamoxifen alone in women older than 50 with node-negative breast cancer.

Table 2. Effect of the addition of RT to t o tamoxifen in women over 50 Tamoxifen (n = 384)

Tamoxifen pl plus RT RT (n = 385)

36

19

IBTR

22 (5.7%)

 p = 0.0009) 2 (0.5%) ( p

Deaths

9

8 (NS)

92.8%

94.7% (NS)

Relapses

OS at 4 years

 

Kuter

 

 Hughess study was simply due to short follow-up and small  Hughe

Table 3. Effect of the addition of RT to t o tamoxifen in women over 70 Tamoxifen (n = 319)

Tamoxifen + RT (n = 317)

In-breast recurrence

4

0

NS

Axillary recurrence

2

0

NS

Eventual mastectomy

1

0

NS

Distant metastasis

1

3

NS

Death, any cause

20

19

NS

Death, breast cancer

1

0

NS

Contralateral breast cancer

4

5

NS

 

340

p

value

In the  Hughes study, 647 women over 70 years old with stage I, estrogen-receptor positive (ER +) breast cancer who were treated with lumpectomy were enrolled and randomly assigned to treatment with tamoxifen plus radiation or tamoxifen alone. At 28 months of follow-up, locoregional recurrences were seen in 6 of 319 women on tamoxifen and 0 of 317 women on tamoxifen plus RT ( p = not significant [NS]) (Table 3). One of 319 women developed distant metastases on tamoxifen versus 3 of 317 on tamoxifen plus RT ( p = NS). There was no effect of radiation on the incidence of ipsilateral breast cancers or on death rate (20 of 319 on tamoxifen; 19 of 317 on tamoxifen plus RT). Only one death was actually cancer related. In this paper, the authors also concluded that when RT is added to tamoxifen, there are fewer locoregional recurrences, but it was pointed out that only one patient required a mastectomy and that contralateral cancers were as common as locoregional recurrences. Furthermore, RT had a detrimental effect on cosmesis. Given the high incidence of deaths from other causes, the low rate of breast recurrence, and the feasibility of breast preservation after in-breast recurrence, the authors raise the possibility that RT may not add much clinical benefit in this older group of women. Commentary Both these studies are being presented at early follow-up (3.4 years in the Fyles study, 28 months in the Hughe  Hughess study). There is no doubt that in both studies, as might be expected, the addition of RT to tamoxifen lowered the locoregional recurrence rate (the lack of statistical significance in the

numbers). The question is whether, in either of the two studies, an argument can be made in favor of tamoxifen without radiation for any subgroup of patients. Further follow-up in the Fyles study is necessary to determine the long-term l ong-term breast control rate following salvage therapy and any differences between the two groups in terms of DFS and OS, but since at 4 years there is already a statistically significant 6% difference in IBTR, it would be hard to recommend lumpectomy and tamoxifen without RT to all women over 50 with nodenegative breast cancer at this time. It is easier to justify the conservative approach of  Hughes in the older women, but even here the death rate (mostly unrelated to breast cancer) is only 6%, so that it might prove hard to justify avoiding radiation on the basis of competing causes of mortality even in this older population of women as a whole. IBTR seems to be less frequent in older women [2], and if patients are infirm or have significant comorbid conditions, the preliminary results of the study can be used to justify withholding radiation. However, at this time even elderly patients outside of a clinical trial should still receive radiation as part of  standard therapy. HORMONAL ISS SSUE UES S

IN

ADJUVANT THERAPY

Bone Mineral Density in Premenopausal Patients in a Randomized Trial of Adjuvant Endocrine Therapy (ZIPP-TRIAL). Á Sverrisdóttir, T Fornander, L

Rutqvist 

(A BSTRACT  96)  96). In this interesting presentation, Sverrisdóttir presented data from the Swedish cohort participating in the European (Sweden/United Kingdom/Italy) ZIPP-TRIAL. Seventy-three premenopausal women with node-negative breast cancer participating in this trial were assigned to adjuvant therapy for 2 years with Zoladex, tamoxifen, Zoladex plus tamoxifen, or no adjuvant endocrine therapy. No adjuvant chemotherapy was given. Total body bone mineral content (TBBM) was measured using dual photon x-ray absorptiometry at initiation i nitiation of treatment, at 12 months, 24 months, and 36 months (one year after treatment ended). As shown in Table 4, there is a significant reduction in bone mineral density after 2 years of   p < 0.0001). In the updated analysis therapy with Zoladex ( p

Table 4. Total body bone mineral content (%) 0 Months

12 Months

24 Months

36 Months

Zoladex

100

96.7

94.8

96.3

Zoladex + Tamoxifen Tamoxifen

100 100

98.9 98.8

99.2 98.4

98.9 97.9

Control

100

98.9

99.2

98.9

 

341

 

presented at the conference, the TBBM actually improved after cessation of Zoladex. Tamoxifen alone did not significantly affect TBBM and compensated for the loss caused by Zoladex. Commentary As we evaluate the various adjuvant options for pre-

menopausal women,Oophorectomy we need to keepmay an eye on the term side effects. indeed be longvery + effective in young women with ER tumors, but the longterm effects of early menopause are concerning. Since some studies using gonadotropin-releasing hormone (GnRH) agonists to suppress ovarian function for only 2 years have given encouraging results, and since, theoretically, the reversible ovarian ablation would have fewer long-term side effects, it is timely to start looking at the “cost” in terms of  bone mineral density, etc., in young women exposed to this form of therapy. Although there are no data yet as to whether the effect of Zoladex leads to any future risk of fractures, in this admittedly small group it is encouraging that although the bone density dropped during the 2 years of  Zoladex, it partly recovered in the following year. Even more interesting is the observation that tamoxifen, which is protective of the bone mineral density in postmenopausal women (but had been feared to decrease bone mineral density in younger women), not only had no significant detrimental effect but also actually seemed to protect the bone mineral density in the women who received Zoladex. This is a very important piece of information as we look at the efficacy and safety of single-hormonal agents compared with combinations in the adjuvant setting.

Breast Cancer

CEF group, 73.9% versus 61.9% ( p = 0.005). However, DIA did not affect relapse-free survival (53% versus 49%) for patients with and without amenorrhea, respectively ( p = 0.3). Commentary Since alkylators have long been thought to be the major etiological agent in drug-induced amenorrhea associated with

adjuvant treatment, it isCMF curious that the CEFcumulative caused a higher of amenorrhea than since doserate of  cyclophosphamide cyclophosph amide was lower after the CEF regimen (median cumulative dose of cyclophosphamide 5,395 mg/m2 in the CEF group compared with 7,839 mg/m2 with CMF). The main point of the presentation, however, is that after stratification for other variables, there was no difference in DFS between those patients who did or those who did not have amenorrhea. The debate over the possible benefit of DIA in women who were premenopausal when breast cancer treatment was initiated has been ongoing for many years. Unfortunately, this study probably does not help settle the issue since it is unlikely that a benefit would be derived unless the amenorrhea is significantly more than 3 months, assuming the mechanism of benefit is analogous to that of tamoxifen (when prolonged antagonism of estrogen is necessary for efficacy). Amenorrhea lasting at least 2 years, or permanent amenorrhea, is more likely to be clinically significant, and the benefit of such can only be assessed in prospective randomized trials, many of which are currently ongoing. Even if  amenorrhea offers some advantage, it has been questioned whether this would afford an increased benefit over tamoxifen alone, and ongoing studies will also help address this issue. None of the women in the above study were given tamoxifen.

Incidence and Prognostic Impact of Amenorrhea During Adjuvant Therapy in High Risk Premenopausal

Preventing Chemotherapy-Associated Amenorrhea

Breast Cancer Patients: Analysis of a National Cancer Institute of Canada Clinical Trials Group (NCIC CTG)

(CRA) with Leuprolide in Young Women with EarlyStage Breast Cancer.  KR Fox, JE Ball, R Mick, HC

Phase III Study. W Parulekar, ME Trudeau, L Shepherd,

 Moore (A BSTRACT  98)  98).

 J Ottaway, A Day, E Franssen, V Bramwell, M Levine,  K Pritchard (A BSTRACT  97)  97). The authors of this study looked at the incidence and prognostic impact of amenorrhea occurring in the National Cancer Institute of Canada Clinical Trials Group study of CEF versus CMF. This study, which demonstrated the superiority of CEF over CMF, involved 716 women of whom 541 fulfilled all the criteria for eligibility, namely normal menstruation at randomization and receipt of six cycles of chemotherapy. Druginduced amenorrhea (DIA) was defined as cessation of menses for ≥3 months during treatment. No patients in the study were given adjuvant tamoxifen. With a median follow-up of 7.7 years, the incidence of DIA was significantly higher in the

Although, as noted above, the benefits of amenorrhea in the young breast cancer patient are still debated, it is not uncommon for a young woman diagnosed with breast cancer to ask for treatment that will optimize her chance of future fertility. In this small study, 13 premenopausal women were treated with leuprolide during their adjuvant chemotherapy with the goal of protecting the ovaries from the cytotoxic effect of chemotherapy. Patients ranged in age from 26 to 39 years old (median 35). Six patients were treated with four cycles of AC, five patients with AC followed by four cycles of paclitaxel, one received CAF × 6 months, and one received doxorubicin/paclitaxel followed by CMF. All were given concurrent leuprolide, and all became amenorrheic by

 

Kuter

the second cycle of chemotherapy; however, all recovered their menses within 1 year of completing chemotherapy.

 

342

Commentary Using ovarian suppression during chemotherapy in an attempt to prevent toxicity to developing follicles is not a new idea, but it has not convincingly been shown to offer an advantage for maintenance of fertility. It is not currently widely used for this reason. In addition, its use during chemotherapy places another burden on the patient, namely coping with acute menopausal symptoms while also dealing with the direct side effects of the chemotherapy. Although, as the author states, it is encouraging that all the patients in this study recovered their periods, it cannot be stated that a role for leuprolide has been demonstrated. Chemotherapy-induced amenorrhea is very age- and regimen-dependent. It has been shown that AC is less likely than CMF to induce amenorrhea, and 11 of 13 patients were treated with this regimen (half also were given Taxol, which, anecdotally, does not seem to be as toxic to the ovary as cyclophosphamide), and the question is whether less ovarian suppression is being seen in this group

P-1 BCPT. In this trial, over 13,000 high-risk women had been randomly assigned to tamoxifen or placebo, and there was a 49% decrease in incidence of breast cancer in the tamoxifen group compared with placebo. When the genetic study began, 315 participants in the BCPT had developed invasive breast cancer. Samples from 288 of these women were available for DNA sequencing, and, of these, 19 (6.6%) had a mutation in either the BRCA1 or BRCA2 gene. Women carrying mutations were more likely to have earlier disease onset (16% of women diagnosed before age 50 had mutations compared with 4% of women diagnosed at age 50 or older) and to have close relatives with breast cancer. Eight women with inherited mutations in BRCA1 BR CA1 developed breast cancer during the BCPT. Five had been randomized to tamoxifen and three to placebo. Thus, the risk ratio for tamoxifen versus placebo among women with BRCA1 mutations was 1.67 (95% CI, 0.41, 8.00) suggesting no reduction in breast cancer incidence. Eleven women with inherited mutations in BRCA2 developed breast cancer. Three had been randomized to tamoxifen and eight to placebo. The risk ratio for tamoxifen versus placebo was 0.38 (95% CI, 0.06,

of young women than would be seen without the use of the leuprolide.  Nanc  Nancyy David Davidson son, in a commentary, quoted the results of a single institution study reported at the San Antonio Breast Conference last year [3] in which only 8% of patients younger than 40 years old treated with AC developed amenorrhea; however, 33% of those treated by AC followed by Taxol did so. At this point, more research needs to be done on the use of GnRH agonists in the adjuvant setting. If there is i s an advantage with respect to future fertility, it will have to be demonstrated in a large prospective randomized trial. There are insufficient data on efficacy to justify use of these drugs for the purpose of maintaining fertility at this time.

1.5). This is equivalent to a reduction in incidence due to tamoxifen of 62%. Although this reduction in incidence was not statistically significant, it is similar to the statistically significant 52% decrease in incidence for the 269 women with detectable mutations in BRCA1 or BRCA2.  Dr. King pointed out that approximately 80% of breast cancers occurring in women with BRCA1 mutations are ER – and this might be why tamoxifen did not reduce the breast cancer risk in BRCA1 mutation carriers. In contrast, 80% of tumors developing in BRCA2 mutation carriers are ER+, and this is probably why tamoxifen was effective in decreasing breast cancer incidence in BRCA2 mutation carriers.

HEMOPREVENT EVENTION ION OF BREAST CAN ANCE CER R IN BRCA1 CHEMOPR AND BRCA2 GENE MUTATION CARRIERS

Commentary These preliminary, long-awaited results are extremely interesting. Given that the work was conducted in the King laboratory and that, during the study, mutations in DNA samples from known mutation carriers not participating in the trial were all identified correctly, it is likely that the results to date are accurate. Although the number of cancers is small, the large size of the study (>13,000 participants) makes it quite unlikely that mutation carriers would be significantly imbalanced between the placebo and tamoxifen groups. Thus the imbalance in occurrence of cancers between the placebo and tamoxifen groups in the BRCA2 mutation carriers, but the lack of imbalance in the BRCA1 carriers, is likely to be meaningful (though not statistically significant). It was shown in the BCPT that the incidence of ER+ tumors was lowered in the participants as a whole by 49%. ER – tumor incidence was, however, not affected by tamoxifen [4]. For the ER + subset,

Tamoxifen and Breast Cancer Incidence Among Women with BRCA1 or BRCA2 Mutations: A Genomics Resequencing Project Embedded in the Breast Cancer Prevention Trial.

In a special session not previously advertised and for which no abstract is available, Mary-Clair  Mary-Clairee King presented, the initial results of protocol P-1G. This is a genomics project embedded in the National Surgical Adjuvant Breast and Bowel Project (NSABP) P-1 Breast Cancer Prevention Trial (BCPT), carried out at the King laboratory at University of  Washington with the collaboration of NSABP investigators. The project used blood samples (anonymized) from patients who developed breast cancer while enrolled on the NSABP

 

 

343

there was a 68% reduction in incidence from tamoxifen treatment. This is similar to the 62% reduction seen from tamoxifen in the BRCA2 mutation carriers, which, as the authors point out, are usually ER+ tumors. It is interesting that not only was there no decrease in risk of cancers in the BRCA1 mutation carriers, there was a trend to increased incidence with the use of tamoxifen. Though clearly not statistically

Breast Cancer

no reason not to assume that tamoxifen will be as helpful in decreasing recurrence risk as it is in other patients with ER+ cancers. TRAST RASTUZUMA UZUMAB B AND HER-2 UPDATE Improved

Survival

Benefit

from

Herceptin

(Trastuzumab) in Patients Selected by Fluorescence in Situ

significant, is  –curious noteseen thatina the trend toward increased incidence ofitER tumorstowas entire group participating in the BCPT trial [4]. A similar trend toward increased incidence of ER – cancers was also seen in the Multiple Outcomes of Raloxifene Evaluation trial in the women treated with raloxifene compared with placebo [5]. These statistically insignificant trends in both trials most likely are a curious coincidence, but, as we prescribe selective estrogen receptor modulators (SERMs) for prevention, it must be borne in mind that we have not totally ruled out a possible increase in incidence of ER – tumors in the treatment group compared with placebo. Are these data persuasive enough for us to prescribe tamoxifen routinely to BRCA2 carriers? Probably most of us would feel comfortable with these data because they are plausible and in line with other data. Should we not prescribe tamoxifen to BRCA1 carriers? This would probably be premature since data from two sources are at odds with these early results from the G-1 study. First of all, oophorectomy is known to protect BRCA1 carriers significantly from developing tumors [6]. Secondly, a large case control study from the Hereditary Breast Cancer Clinical Study Group [7] showed that tamoxifen use decreased the incidence of contralateral breast cancers in BRCA1 carriers by 62% and in BRCA2 carriers by 32%. The protective effect was independent of that of oophorectomy. One explanation of the apparent discrepancy in the results of  these studies is that only 13% of the BRCA1 carriers in the

Mass, M Pres Mass, Press, s, S Ande Anderso rson, n,  BSTRACT  ACT 85).  M Murphy, Murphy, D Slamo Slamon n (A BSTR

 Narod study had received tamoxifen. It is likely that tamox-

group (38% versus 37.5%,  p = NS). FISH results also predicted survival benefit: in the FISH+ group, addition of  Herceptin to chemotherapy provided a significant survival benefit (OR 0.71,  p = 0.009) not seen in the FISH – group (OR 1.11, p = NS).  Dr. Vogel Vogel presented data on FISH analysis and response to single-agent single-agent Herceptin Herceptin from two different different trials. In trial

ifen had only been prescribed for the BRCA1 carriers who had ER+ tumors. If patients with ER+ tumors are at increased risk of having a second ER+ tumor, the protective benefit would be explained. Finally, the authors took pains to point out that their result pertained only to prevention. If a patient with a BRCA1 mutation develops ER+ breast cancer, there is

Hybridization (FISH).  RD

Superior Outcomes with Herceptin (Trastuzumab) (H) in Fluorescence in Situ Hybridization (FISH)-

CL Vogel, M Cobleigh, D Tripathy,  R Mass, M Murphy, SJ Stewart (A BSTRACT  86). Selected Patients.

In the pivotal phase III trial of chemotherapy (AC or Taxol) ± trastuzumab as first-line treatment treatment of HER-2-overexpressing breast cancer, there was an advantage to the addition of Herceptin in terms of response rates (50% versus 38%) and survival (odds ratio [OR], 0.80) despite a trial design resulting in 65% of control patients receiving Herceptin at disease progression [8]. To be eligible for this trial, patients’ tumors had to overexpress HER-2 (2+ or 3 + by IHC). The IHC was run in a central lab. In the current study, archived blocks from patients in the study were sub jectedd to FISH analy  jecte analysis. sis. A HER-2 HER-2:CEP :CEP 17 sign signal al ratio was detected using the PathVysion dual probe FISH assay system. HER-2 gene amplification was detected in 76% of the patients on the study. Ninety-two percent of 3 + overexpressors and 32% of 2 + overexpressors were FISH+. The addition of Herceptin to chemotherapy improved the response rates in the FISH+ subgroup (54% versus 30.8%, p < 0.0001) (Table 5), while there was no improvement in the FISH – sub-

Table 5. Treatment outcomes in patients treated with chemotherapy chemotherapy ± Herceptin Response Rate

Overall Survival

C

C+H

C

C+H

FISH –

37.5%

38%

19.8 months

24 months NS

+

30.8%

54%

20 months

26.2 months sig

FISH

Abbreviations: C = chemotherapy; H = Herceptin

 

Kuter

 

344

Table 6. Treatment outcomes in Herceptin single-agent trials H0650g

H0649g

RR (%)

CR+PR+SD (%)

OS (months)

RR (%)

CR+PR+SD (%)

OS (months)

34 (24, 45)

48 (37, 60)

24.5 (17.4, 36.1)

19 (14, 26)

33 (27, 41)

14.2 (12.4, 18.1)

FISH –

7 (1, 23)

10 (2, 27)

24.4 (10.8, 34.7)

0 (0.0, 9.7)

6 (1, 19)

8.8 (5.8, 15.6)

IHC+

26 (19, 36)

38 (29, 48)

24.4 (16.9, 31.7)

15 (11, 21)

28 (23, 35)

12.8 (11.5, 15.7)

FISH

+

H0649g, patients were offered single-agent H0649g, single-agent Herceptin after one or two chemotherapy regimens for metastatic disease (including an anthracycline and a taxane). In H0650g, Herceptin was offered as first-line single-agent treatment for metastatic disease. As in the pivotal trial, eligibility was assessed initially by IHC. FISH showed HER-2 gene amplification in 92% of 3 + overexpressors and 39% of 2 + overexpressors. As shown in Table 6, FISH positivity predicted for a higher response rate to the Herceptin in both studies. For all patients in H0649g, the response rate was 15%, but for those patients whose tumors were FISH –, the response rate was 0% while for those with FISH + tumors, it was 19%. Clinical benefit rates (complete response [CR] + partial response [PR] + stable disease [SD] for >6 months) were 28% overall, but only 6% in the FISH – subgroup compared with 33% in the FISH + subgroup. For the H0650g trial (first-line treatment), response was 26% overall but only 7% in the FISH – group compared with 34% in the FISH + group. Clinical benefit rate in this trial was 38% (10% of  FISH –, 48% of FISH+). The conclusion was that FISH is superior to IHC in selecting patients for Herceptin treatment and that monotherapy with Herceptin can be a potent treatment for HER-2+ metastatic breast cancer, particularly

used, should be the test of choice. Recently the use of FISH has been emerging as the more reproducible and dependable test. In the FISH test, the relative amounts of the HER2 gene are expressed as a function of the amount of  centromeric signal from the same chromosome (no. 17). Amplification of the HER-2 gene is present if multiple copies of the HER-2 gene are noted (ratio >2). The new data presented at the conference by these investigators showed that the response to Herceptin occurs predominantly in patients whose tumors are positive by FISH, confirming that the FISH assay is more reliable than IHC in identifying candidates for Herceptin. Given the higher cost of the FISH assay, however, it is currently recommended that all samples are initially screened by IHC. Patients with 3+ IHC staining are candidates for Herceptin treatment. Patients with 2+ IHC staining should have their tumors tested by FISH to distinguish false positives from true overexpression. Furthermore, any tumor that is high grade and ER – might be screened by FISH even if the IHC is negative to ensure that the IHC results are not falsely negative since it is predicted, based on correlative studies, that about 10%15% of tumors with amplified genes will be read as negative by IHC. A word of caution: anecdotally, there have

for first-line treatment.

been a number of cases in which the tumor appeared to be HER-2+ on IHC based on staining of a core biopsy, but the HER-2 has proved to be negative on a subsequently excised specimen. It is possible that false positives occur because of  crush artifact in the core specimen. FISH might be preferable to IHC on cores if Herceptin is being considered in the neoadjuvant setting. In light of the data from H0650g in which a clinical benefit rate of 48% was seen in FISH + patients, it is clear that monotherapy with Herceptin in appropriately selected patients with metastatic disease is very reasonable. As  Dr. Vogel suggested, one might apply the same reasoning to selection of patients for treatment with Herceptin alone as one might use in selection for hor-

Commentary The two pivotal trials that led to the Food and Drug Administration approval of Herceptin were the H0649g trial of monotherapy with Herceptin showing a 15% response rate to Herceptin in previously treated patients [9] and the phase phase III of chemotherapy chemotherapy ± Herceptin as first-line treatment of HER-2-overexpressing metastatic breast cancer, which demonstrated superiority of the chemotherapy/Herceptin combination over chemotherapy alone [8]. For both these trials, eligibility was restricted to those women whose tumors overexpressed HER-2 by IHC (2+ or +

3 overexpressing, about 30% of breast cancers). Since then, there has been a raging debate about whether IHC, the results of which are quite subjective (at least if the staining is less than 3+) and can depend on the antibody and conditions

monal therapy, i.e., patients without life-threatening organ involvement who can afford to wait a couple of months to assess the possible response before being subjected to chemotherapy.

 

 

345

TAMOXIFEN RESISTANCE Effect of Raloxifene After Tamoxifen on Breast and Endometrial Cancer Growth.  RM

O’Regan, C Gajdos,  R Dardes, A de los Reyes, DJ Bentrem, VC Jordan (A BSTRACT  95)  95).

Breast Cancer

growth. At present, the American Society of Clinical Oncology guidelines [10] caution that there are no good data to support the sequential use of SERMs, such as tamoxifen followed by raloxifene. The EGFR-Selective Tyrosine Kinase Inhibitor ZD1839 (Iressa) is an Effective Inhibitor of Tamoxifen-

O’Regan et al this interesting study, . looked at animal In cancer models with respect to the effect of raloxifene given after tamoxifen. In each model, human tumor cells were grown in athymic mice in the presence of estrogen, tamoxifen, raloxifene, or a SERM together with estrogen. The mice were implanted on one side with breast cancer cells and on the other side with endometrial cancer cells. When the breast cancer cells were from a line resistant to tamoxifen (generated by long-term exposure to the drug), tamoxifen actually stimulated the growth of the tumor in the athymic mice, and raloxifene stimulated these cells to the same degree. When the source of the endometrial cancer cells was a tamoxifen-resistant line, again there was stimulation of growth by tamoxifen and a similar degree of  stimulation by raloxifene, suggesting tamoxifen resistance in both a breast cancer and an endometrial cancer model confers raloxifene resistance, too. Commentary Although not a clinical study, this intriguing report from  Dr. Jordan’s Jordan’s group merits contemplation. In brief, raloxifene is considered to be a safer drug for long-term use than tamoxifen, since it appears not to stimulate the endometrium. After 5 years of adjuvant tamoxifen for breast cancer, many physicians are putting their patients on raloxifene for preservation of bone mineral density. Tamoxifen is not currently recommended for more than 5 years in the adjuvant setting because

in two small studies randomizing women to 5 years versus 10 years of tamoxifen, there were actually more recurrences in the group assigned to 10 years, and there was also an increase in endometrial cancer risk in the patients given 10 years of the drug. The question arises as to whether substituting raloxifene after 5 years of tamoxifen is safer than continuing tamoxifen for 10 years. This study suggests that it may, in fact, not be a good idea since raloxifene was unable to inhibit tamoxifenresistant breast cancer cells and, in fact, appears to stimulate their growth in a fashion similar to tamoxifen. Furthermore, raloxifene, tamoxifen, and estrogen had similar stimulating effects on the growth of the endometrial cancer cells. Until clinical trials provide more data, the possible benefit of raloxifene after 5 years of tamoxifen needs to be balanced against the possible risks of stimulating dormant residual cancer cells and even possibly the risk of acting with the prior tamoxifen to stimulate endometrial cancer

Resistant Breast Cancer Growth. JM Gee, IR IR Hutcheson, Hutcheson,

 JM Knowlden, D Barrow, ME Harper, AE Wakeling,  RI Nicholson (A BSTRACT  282)  282). In another study using a breast cancer cell line model MCF7, breast cancer cells exposed to tamoxifen acquired resistance to the drug. These cells were found to overexpress EGFR and HER-2, which were present as heterodimers. Tumor necrosis factor-α, an epidermal growth factor receptor (EGFR) ligand, increased the phosphorylation of the EGFR and the growth of the tamoxifen-resistant cells to a much greater degree than wild-type cells. ZD1839 (Iressa), a selective EGFR small molecule inhibitor, was able to wipe out the phosphorylation of the EGFR and of phosphorylated intermediates in the MAP kinase pathway, resulting in marked inhibition of growth of the tamoxifen-resistant cells. In contrast, the drug had no effect on wild-type (tamoxifen-sensitive) cells. When wild-type cells were grown in the presence of tamoxifen, growth inhibition by tamoxifen gradually declined, resulting resul ting in resistance r esistance by 3 months. months. The The acquisition of resistance was abolished if Iressa was also added to the cells with the tamoxifen. The author concludes that the EGFR/MAP kinase pathway is important in the generation of tamoxifen resistance in this model system, and its inhibition by Iressa could prevent the development of this resistance in vitro. Commentary Although tamoxifen is a potent drug for treating ER + breast cancer, drug resistance usually develops after 1 to 2 years of treatment. The EGFR is also commonly overexpressed in breast cancer, and it has been suggested that growth stimulation via the EGFR pathway might be an important mechanism by which tamoxifen resistance can occur. HER-2 overexpression has also been linked to tamoxifen resistance. In this study, data are presented to support the hypothesis that tamoxifen resistance occurs because an alternative pathway of growth stimulation, via the MAP kinase pathway (stimulated by an activated EGFR) stimulates growth of the cells. Iressa was able to reverse the molecular activation of the MAP kinase pathway seen in the tamoxifen-resistant cells, and this was accompanied by effective growth inhibition of tamoxifenresistant (but not tamoxifen-sensitive) cells. Although a

 

Kuter

cautionary note should be added that data from cells lines can differ significantly from real life cancers, this exciting observation suggests that the drug be tested in clinical trials to see if it can prevent or reverse resistance to tamoxifen in patients. It should be noted that studies are already under way to see if Herceptin can restore tamoxifen sensitivity to ER+ breast cancers overexpressing HER-2 since HER-2 overexpression also leads to a relative resistance to antiestrogen

 

346

treatment. Studies of the interaction of the tyrosine kinasemediated pathways and the ER pathways might lead to some exciting new developments in endocrine treatment of  breast cancer in the next few years. ACKNOWLEDGMENT  I.K. is a member of the speakers’ bureau for AstraZeneca and Genentech.

R EFERENCES 1 Dowlatshahi K, Fan M, Snider Snider HC et et al. Lymph node micromicrometastases from breast carcinoma: reviewing the dilemma. Cancer 1997;80:1188-1197. 2 Nixon AJ, Neuberg Neuberg D, Hayes Hayes DF et al. Relationship Relationship of patient patient age to pathologic features of the tumor and prognosis for patients with stage I or II breast cancer. J Clin Oncol 1994;12:888-894. 3 Stone ER, Slack Slack RS, Novielli Novielli A et al. Rate of chemotherapy chemotherapy related amenorrhea associated with adjuvant Adriamycin and Cytoxan (AC) and Adriamycin and Cytoxan followed by Taxol (AC+T) in early stage breast cancer. 23rd Annual San Antonio Breast Cancer Symposium. Breast Cancer Res Treat 2000;64:61a. 4 Fisher B, Costantino JP, Wickerham Wickerham DL et al. Tamoxifen Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast and Bowel Project P-1 Study. J Natl Cancer Inst 1998;90:1371-1388. 5 Cauley JA, Norton Norton L, Lippman ME et al. Continued breast cancer risk reduction in postmenopausal women treated with

raloxifene: 4-year results from the MORE trial. Breast Cancer Res Treat 2001;65:125-134. 6 Rebbeck TR, Levin AM, Eisen A et al. Breast Breast cancer risk after bilateral prophylactic oophorectomy in BRCA1 mutation carriers. J Natl Cancer Inst 1999;91:1475-1479. 7 Narod SA, Brunet JS, Ghadiria Ghadiriann P et al. Tamoxifen Tamoxifen and risk risk of  contralateral breast cancer in BRCA1 and BRCA2 mutation carriers: a case-control study. Lancet 2000;356:1876-1881. 8 Slamon DJ, Leyland-Jones Leyland-Jones B, Shak S et al. Use of chemotherapy chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N Engl J Med 2001;344:783-792. 9 Vogel CL, Nabholtz JM. Monotherapy of metastatic breast cancer: a review of newer agents. The Oncologist  1999;4:17-33. 10 Chlebowski RT, Collyar DE, Somerfield Somerfield MR et al. al. American Society of Clinical Oncology technology assessment on breast cancer risk reduction strategies: tamoxifen and raloxifene. J Clin Oncol 1999;17:1939-1955.