[CANCER RESEARCH 5|, 944—948. February 1. 1991]
`
`Follow-up Study of HER-Z/neu Amplification in Primary Breast Cancer1
`
`Gary M. Clark2 and William L. McGuire
`Department ofMedicine/Medical Oncology, University of Texas Health Science Center at San Antonio, San Antonio, Texas 78284— 7884
`
`ABSTRACT
`
`Amplification of the HER-Z/neu oncogene was determined in 362
`tumors from patients with primary breast cancer (185 node-positive
`patients and 177 node-negative patients). The overall amplification rate
`was 33% (30% for node-negative patients; 31% for patients with 1—3
`positive nodes; 40% for patients with >3 positive nodes). Gene copy
`number was not associated with axillary lymph node status, steroid
`receptor status, or patient age but was weakly correlated with the size of
`the primary tumor. Amplification of the HER-Z/neu gene did not correlate
`with either disease-free or overall survival in univariate or multivariate
`analyses. The results were unambiguously negative for patients with
`node-negative disease. Although the univariate results for node~positive
`patients were marginally significant (P = 0.07), the significance was not
`retained in multivariate analyses. Thus, while HER-Zlneu amplification
`may be biolofically important in primary breast cancer, it will only be of
`marginal utility as a prognostic factor for predicting clinical outcome.
`
`INTRODUCTION
`
`The HER—Z/neu oncogene has been the subject of heated
`debates concerning its prognostic significance for women with
`breast cancer. In our initial report, we found that the HER-2/
`neu gene was amplified in 28% of the 189 primary breast tumors
`examined (1). This amplification was directly related to the
`number of axillary lymph nodes that were involved with tumor
`but was unrelated to other prognostic factors including steroid
`receptor status, tumor size. and patient age. The most exciting
`finding was that HER—Z/neu amplification was an important
`predictor of disease recurrence and death for 86 patients with
`axillary node-positive disease.
`Several reports have now been published by other investiga—
`tors who have also studied HER—Z/neu amplification in human
`primary breast tumors (2-26). The reported amplification rates
`range from 10 to 46% (Table l), with an overall amplification
`rate of 20% based on 2992 patients. In general, patients with
`node-positive disease have greater frequency of amplification
`than patients with node-negative disease. However, there has
`been considerable controversy regarding the correlations be—
`tween HER-Z/neu amplification and clinical outcome. Most of
`these reports have described relationships between gene ampli-
`fication and other prognostic factors but have not been able to
`correlate amplification with disease outcome because of small
`numbers of patients with short follow—up times. Of those studies
`that have examined clinical correlations, six have reported a
`direct relationship between HER-Z/neu amplification and poor
`survival (1, 3, 14-16, 23), and three studies have claimed that
`gene amplification is not related to patient outcome (9, 12, 24).
`An additional study (23) observed a significant relationship
`with early relapse and death in univariate analyses that disap-
`peared in multivariate analyses. Several possible explanations
`
`Received 3/5/90; accepted 11/20/90.
`The costs of publication of this article were defrayed in part by the payment
`of page charges. This article must therefore be hereby marked advertisement in
`accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
`' This work was supported in pan by NIH Grant CA 30195.
`1 To whom reprint requests should be addressed. at Department of Medical
`Oncology, University of Texas Health Science Center at San Antonio, 7703 Floyd
`Curl Drive, San Antonio, TX 78284-7884.
`
`‘°"
`944
`
`NAHW‘IAAAAIJ ‘V‘AM u'luu'ludIudH-J'LV‘IJAA AAASIA.W-nalh At”. A“ finanmkny (2
`
`for low amplification frequency and lack of clinical correlations
`have been proposed (10), including inadequate analytical tech-
`niques, suboptimal statistical analyses, and small numbers of
`patients.
`In this study, we have expanded our previous study design to
`include 185 additional node-positive patients and 177 node—
`negative patients in order to examine the prognostic ability of
`HER-Z/neu amplification for patients with primary breast can—
`cer.
`
`MATERIALS AND METHODS
`
`Patients. The tumor specimens used in this study were nuclear pellets
`selected from our San Antonio Tumor Bank, which has been described
`previously (27). These tumors were originally sent to our laboratory for
`determination of steroid receptor content. After completion of the
`receptor assays, the remaining nuclear pellets were stored in freezers
`(—70‘C) to be used in future studies. Demographic characteristics,
`treatment information, and follow-up status of each patient for disease
`recurrence and mortality are routinely collected by a specialized team
`of data managers.
`The eligibility criteria for inclusion in this study were a tumor
`specimen obtained between 1973 and 1985 at the time of diagnosis
`from a woman with primary breast cancer, evaluable steroid receptor
`results, documented tumor size, and age of the patient at the time of
`diagnosis. The study was designed to yield approximately 400 evaluable
`specimens for HER-Z/neu amplification analysis assuming a 60% eval-
`uability rate.
`Oncogene Amplification. HER—Z/neu amplification was determined
`in a single laboratory by Southern analysis as previously described (1,
`10, 15) in a blinded fashion without knowledge of the patient’s clinical
`outcome or tumor characteristics. In brief, DNA from the individual
`tumors was digested with EcoRI and subjected to Southern blot analysis
`with a 32Illabeled HER-Z/neu-l probe (provided by Axel Ullrich) which
`is known to detect a lS-kilobase hybridizing band in human DNA. All
`DNA blots were stripped and reprobed with both p53 and myeloper—
`oxidase probes to evaluate the relative loading of DNA in each lane
`and to exclude the possibility that amplification was caused by partial
`or complete duplication of chromosome 17. Blots were scanned by soft
`laser densitometry and the level of HER-Z/neu amplification was
`determined by the ratio of the HER—Z/neu signal relative to the single-
`copy p53 signal.
`Statistical Analyses. The data were maintained in an AT&T 3815
`minicomputer using the Informix data base management system (In-
`fonnix Software, Menlo Park, CA). Statistical analyses were performed
`with the BMDP statistical package (BMDP Statistical Software, Los
`Angeles, CA). For statistical analyses, HER-Z/neu gene amplification
`was coded as: 1, single copy; 2, 2-5 copies; 5, 5—20 copies; 20, >20
`copies. Other prognostic factors were dichotomized as steroid receptor
`negative versus steroid receptor positive (using 3 and 5 fmol/mg protein
`for ER and PgR, respectively), the number of positive axillary lymph
`nodes (1—3 versus >3), tumor size (52 versus >2 cm), and age (550
`versus >50 years).3 Correlations between HER-Z/neu amplification and
`other prognostic factors were evaluated using )8 tests for trends. The
`primary end points in this study were disease-free survival and overall
`survival. Disease-free survival was defined as the interval between the
`diagnostic biopsy and the first recurrence of breast cancer. Patients
`who died without documented disease recurrence were considered cen—
`
`’ The abbreviations used are: ER, estrogen receptor, PgR, progesterone rwep—
`
`Genentech 2095
`
`Hospira v. Genentech
`.
`|PR2017-OO737
`find—l KA 4004 AMAv-iAAw AAAAAC AAAAAA
`
`Genentech 2095
`Hospira v. Genentech
`IPR2017-00737
`
`1
`
`

`

`Table l Amplification of HERZ/neu in primary breast cancer
`Patients
`
`
`Ref.
`All
`Node-positive Node-negative
`Slamon et al. (1)
`53/189 (28)‘
`44/133 (33)
`4/34 (12)
`van de Vijver er al. (2)
`15/95 (16)
`6/41 (15)
`6/44 (14)
`Varley er a1. (3)
`7/41 (17)
`6/25 (24)
`0/9 (0)
`Venter er al. (4)
`12/36 (33)
`3/15 (20)
`4/16 (25)
`Cline et al. (5)
`8/53 (15)
`8/35 (23)
`0/14 (0)
`Zhou et al. (6)
`15/86 (17)
`8/37 (22)
`1/21 (5)
`Berger et al. (7)
`13/51 (25)
`7/17 (41)
`5/26 (19)
`Tal et al. (8)
`2/21 (10)
`Ali era]. (9)
`12/122(10)
`Fontaine era]. (11)
`7/15 (46)
`Zhou eral.(12)
`17/157(11)
`Zeillinger er a1. (13)
`52/291 (18)
`Tsuda etal.(14)
`28/176(16)
`Slamon er al. (15)
`146/526 (28)
`King et al. (16)
`12/61 (20)
`R0 er al. (17)
`13/66 (20)
`Adnane er al. (18)
`45/219 (21)
`Garcia er al. (19)
`27/125 (22)
`LaCroix er al. (20)
`5/42 (12)
`Tavassoli er al. (21)
`15/52 (29)
`Yamada et al. (22)
`10/50 (20)
`Borg er al. (23)
`42/279(15)
`Heintz er a1. (24)
`17/50 (34)
`Hanna er al. (25)
`16/66 (24)
`Uehara er al. (26)
`19/123 (15)
`
`8/75(11)
`5/11 (45)
`10/89(11)
`
`20/104(l9)
`101/345 (29)
`4/26 (15)
`
`18/98 (18)
`12/55 (22)
`2/30 (7)
`11/25 (44)
`10/31 (32)
`22/120(18)
`10/29 (34)
`
`14/76 (18)
`
`0/1 (0)
`6/68(9)
`
`8/72(11)
`45/181 (25)
`8/35 (23)
`13/66 (20)
`25/111(23)
`14/66 (21)
`3/12 (25)
`4/27 (15)
`0/19 (0)
`20/159(l3)
`7/21 (33)
`16/66 (24)
`5/47 (11)
`
`HER-Z/uu AMPLlFlCATlON 1N BREAST CANCER
`Table 2 Comparison between evaluable and unevaluable patients
`
`
`
` Factor Evaluable Unevaluable P value
`Positive nodes
`0
`1—3
`4+
`ER status
`—
`+
`PgR status
`—
`+
`Tumor size (cm)
`52
`>2
`Age (yr)
`550
`>50
`Chemotherapy
`No
`Yes
`Endocrine therapy
`No
`Yes
`5-yr disease-free survival
`5-y'r overall survival
`' Numbers in parentheses, %.
`' Actuarial estimate 1 SD.
`
`177 (49)“
`99 (27)
`86 (24)
`
`92 (25)
`270 (75)
`
`163 (45)
`199 (55)
`
`126 (35)
`236 (65)
`
`155 (32)
`247 (68)
`
`246 (70)
`107 (30)
`
`146 (49)
`86 (29)
`68 (23)
`
`83 (28)
`217 (72)
`
`145 (48)
`155 (52)
`
`116 (39)
`182 (61)
`
`79 (36)
`221 (74)
`
`218 (73)
`82 (27)
`
`0.91
`
`0.51
`
`0.40
`
`0.27
`
`0.13
`
`0.40
`
`238 (79)
`250 (71)
`62 (21)
`104 (29)
`0.21
`as 1 3%
`74 x 3%‘
`88 :1: 2% 0.23 83 t 2%
`
`
`
`0.01
`
`Table 3 HER-2/neu gene copy number by prognosticfactor
`
`Copies
`
`Factor
`Single
`2—5
`5—20
`>208 Total
`Positive nodes (P = 0.42)“
`0
`1—3
`4+
`ER status (P = 0.24)
`—
`+
`PgR Status (P = 0.38)
`—
`+
`Tumor size (cm) (P = 0.06)
`52
`>2
`Age (yr) (P = 0.29)
`550
`>50
`Chemotherapy (P = 0.92)
`No
`Yes
`Endocrine therapy (P = 0.48)
`250
`14(6)
`45 (18) 18(7)
`173 (69)
`No
`Yes 104 63 (61) 23 (22) 13 (12) 5(5)
`
`
`
`
`
`‘ P values by x1 test for trend.
`" Numbers in parentheses. %.
`
`194/1115 (17)
`329/1417 (23)
`608/2992 (20)
`Total
`‘ Number of patients/total number of patients. Numbers in parentheses, %.
`The numbers for all patients do not necessarily equal the sum of node-positive
`and node—negative patients because nodal status was not known for all patients.
`
`sored for disease-free survival but were included as deaths when analyz—
`ing overall survival. Curves for disease-free and overall survival were
`calculated according to the method of Kaplan and Meier (28). The
`differences between curves were assessed with the log-rank test for
`censored survival data (29). The partially nonparametric regression
`model of Cox (30) was used to evaluate the predictive power of various
`combinations of prognostic factors in a multivariate manner.
`
`RESULTS
`
`123 (70)” 34 (19) 13(7)
`68 (69)
`18 (18)
`6(6)
`52 (60)
`17 (20)
`12 (14)
`
`7(4)
`7 (7)
`5(6)
`
`65 (71)
`178 (66)
`
`9 (10)
`5(5)
`13 (14)
`56 (21) 26 (10) 10(4)
`
`111(68)
`132 (66)
`
`14 (9)
`27 (17)
`42 (21) 17(9)
`
`85 (68)
`158 (67)
`
`11 (9)
`28 (22)
`41 (17) 20(9)
`
`11(7)
`8(4)
`
`2 (2)
`17(7)
`
`177
`99
`86
`
`92
`270
`
`163
`199
`
`126
`236
`
`87 (76)
`156 (63)
`
`9 (8)
`14 (12)
`55 (22) 22 (9)
`
`5(4)
`14 (6)
`
`158 (64)
`77 (72)
`
`52 (21) 25 (10) 11(5)
`16 (15)
`6(6)
`8(8)
`
`115
`247
`
`246
`107
`
`A total of 662 frozen nuclear pellets obtained after perform-
`ing steroid receptor assays on frozen biopsy specimens from
`patients with primary breast cancer were selected for inclusion
`in this study (323 from patients with node-negative breast
`cancer and 339 with node-positive disease). DNA was success-
`fully extracted from 362 (55%) of these samples. The primary
`reasons for failure to obtain DNA were (a) too few viable cells,
`and (b) degradation of the DNA that was extracted. There were
`no significant differences between patients with evaluable HER—
`larger tumors to have increased numbers of copies of the HER-
`2/neu gene (P = 0.06).
`2/neu results and patients without results with respect to num-
`Adjuvant systemic therapy (chemotherapy plus or minus
`ber of positive axillary lymph nodes, ER status, PgR status,
`endocrine therapy) was administered to 78% of the node-
`tumor size, and the administration of chemotherapy (Table 2).
`Slightly more patients with evaluable results received endocrine
`positive patients but to only 20% of node-negative patients.
`therapy (29%) compared to patients without results (21%) (P
`There was a tendency (P = 0.10) for node-positive patients who
`= 0.01), but this did not translate into significant differences
`received adjuvant endocrine therapy to have had an amplified
`between the two groups of patients with respect
`to either
`HER-Z/neu gene (39%) compared to untreated patients (31%),
`disease-free or overall survival.
`but administration of chemotherapy was unrelated to amplifi-
`The overall HER—Z/neu amplification rate for the 362 eval-
`cation of the gene. In contrast, only 15% of node-negative
`uable patients was 33% (30% for node-negative patients; 31%
`patients who received adjuvant chemotherapy had an amplified
`for patients with 1—3 positive nodes; 40% for patients with >3
`gene compared to 33% who were not treated (P = 0.055).
`positive nodes). This compares favorably with the results in our
`A total of 362 patients had evaluable HER-Z/neu results. The
`originalgroup (28% for patients with 1—3 positive nodes and
`median follow—up of these patients was 75 months (83 months
`51% for patients with >3 positive nodes). HER-Z/neu amplifi-
`for patients still alive at the time of analysis). Disease-free and
`cation was not associated with lymph node status, ER status,
`overall survival are displayed in Fig.
`1 and 2, respectively, by
`PgR status, age of the patient, or administration of adjuvant
`the number of copies of the HER-Z/neu gene. Log-rank tests
`endocrine or chemotherapy (Table 3). There was a trend for
`comparing the four groups of patients (single copy, 2—5 copies,
`945
`
`hat-.mIAnAAA [VA-ev- nnnnnnnnnnnnnn'. ...... I ~~~~~~ nannmkav {2 ”H47 R 4nn4 AM nnnnnn AnnAA:n‘:r\v~ 5"“, f" AAAAA
`
`2
`
`

`

`HERvZ/neu AMPLIFICATION 1N BREAST CANCER
`
`:J-rr—x,
`
`'_‘H«__
`1
`1.. 31:11 1_
`"
`-~_1
`:
`'L__
`
`2-5 Copies (ii-69)
`-» ———L
`_ H _
`'SingleCopy
`— :— (fl-243)
`5-20 Copies (n-31)
`
`> 20 Copies (n-19)
`
`0‘3 '
`>~
`-
`:
`i 03.1
`g
`I
`i 0.4
`as
`
`.
`
`02
`
` -
`‘1
`12 24364860728496
`
`o
`
`Time (months)
`Fig. 1. Disease-free survival (DFS) by number of copies of HER-2/neu gene.
`Amplification of HER—Z/neu did not correlate with disease~free survival for 362
`patients with median follow-up of 75 months. The P values for log—rank tests
`were: 0.15 when the four groups were compared; 0.74 when single copy was
`compared to 22 copies; 0.21 when <5 copies were compared to 25 copies; 0.07
`when copy number was coded I, 2. 5. or 20 as a continuous factor; and 0.42 when
`copy number was coded l. 2, 3. or 4 as a continuous factor.
`
`2-5 Copies (n-69) - —L—- 5-20 Copies
`
`' "--;,_
`Sin 1e Co
`g
`
`(fl-31)
`(n-243)
`”V
`
`2.
`.__
`(O
`5 0.6
`.t:
`
`02 1
`
`Consistent with the results for the combined analyses, HER-
`2/neu amplification was not a significant predictor of clinical
`outcome for the 177 patients with node-negative disease, either
`as a single factor or in multivariate analyses (Fig. 3). In fact,
`node—negative patients with amplified HER-Z/neu had slightly
`better clinical outcomes than patients with a single copy of the
`gene. Similarly, gene amplification was not predictive of patient
`outcome in the 185 node-positive patients (Fig. 3). When copy
`number was analyzed as a continuous factor, a marginally
`significant trend was observed between gene copy number and
`disease-free survival (P = 0.11), but no relationship was found
`for overall survival (P = 0.39). In multivariate analyses, the
`number of positive nodes was the most important predictor of
`disease-free survival (P = 0.0042), and HER-Z/neu amplifica-
`tion remained only marginally significant (P = 0.10). ER status
`and the number of positive nodes were the only significant
`predictors of overall survival. This is in contrast to our original
`group of 86 node-positive patients in which HER-Z/neu ampli—
`fication was the most important predictor of both disease-free
`and overall survival.
`
`DISCUSSION
`
`The HER-Z/neu gene was amplified in 33% of our primary
`breast tumors. This is consistent with our previous series of
`patients (1) but is slightly higher than many of the reports in
`
`Univariate Multivariate
`Univariate Multivariate
`
`
` Factor P value P value RR” P value P value RR
`
`
`Lymph node status
`0.0035
`0.008
`1.66
`0.075
`0.18
`(1—3 vs. >3)
`PgR status (positive
`vs. negative)
`Tumor size (52 vs.
`>2 cm)
`HER-Z/neu (1. 1; 2,
`2—5; 5, 5—20; 20,
`>20)
`Age (550 vs. >50
`yr)
`0.85
`0.65
`0.48
`0.64
`ER status (positive
`vs. negative)
`‘ Follow-up median (range): all patients, 71 months (1—163); still alive, 83
`months (1—163).
`" RR. relative risk defined as last category relative to first category.
`
`0.024
`
`0.016
`
`0.073
`
`1.53
`
`1.43
`
`0.020
`
`0.081
`
`0.20
`
`0.48
`
`0.44
`
`0.49
`
`0.006
`
`0.48
`
`0.67
`
`
`
`0.58
`
`0.007
`
`1.92
`
`0.77
`
`0.64
`
`10 'r
`
`N-, Ampifred (11'64) 2:
`
`u.-
`
`N-. Singe Copy 01-123)
`
`2-
`
`g on
`'8
`6‘:
`Qo
`
`0.4
`
`02 J
`
`_
`
`*—-_.,__ _ N+. Single Copy
`"'"1—1
`(rt-120)
`N+. Amplified (rt-65)
`
`-.—j~—~._-s
`r
`o.oj——r
`01224364860728496
`Time (months)
`
`0.4 -
`
`g
`U)
`0
`
`> 20 Copies (11-19)
`
`Table 4 Univariate and multivariate analyses ofdisease-free and overall survival
`
`for 362 patients”
`Disease-free survival
`
`Overall survival
`
`00 fi—r-‘wfifi—' ' T _T‘—"7
`O
`12
`24
`36
`48
`60
`72
`84
`96
`
`Time (months)
`Fig. 2. Overall survival (OS) by number of copies of HER-Z/neu gene. Ampli-
`fication of HER-Z/neu did not correlate with overall survival for 362 patients.
`The P values for log-rank tests were: 0.37 when the four groups were compared;
`0.33 when single copy was compared to 22 copies; 0.90 when <5 copies were
`compared to 25 copies; 0.48 when copy number was coded l, 2, 5. or 20 as a
`continuous factor; and 0.75 when copy number was coded l. 2, 3. or 4 as a
`continuous factor.
`
`5—20 copies, >20 copies) indicate that the gene copy number
`did not correlate with either disease-free or overall survival (P
`= 0.15 and P = 0.37, respectively). Similar results were observed
`using other representations of amplification to predict disease-
`free or overall survival: single copy versus multiple copies (P =
`0.74 and P = 0.33); <5 copies versus >5 copies (P = 0.21 and
`P = 0.90); copy number as a continuous factor coded 1, 2, 5,
`or 20 (P = 0.07 and P = 0.48); copy number as a continuous
`factor coded 1, 2, 3, or 4 (P = 0.42 and P = 0.75).
`Multivariate analyses (Table 4) revealed that lymph node
`status, PgR status, and tumor size predicted disease recurrence,
`but only tumor size was statistically significant for predicting
`overall survival. After adjustment for other prognostic factors,
`HER—Z/neu amplification was not a significant predictor of
`either disease-free or overall survival regardless of how gene
`amplification was represented in the models.
`Since the only significant correlations that have been de—
`scribed in the literature between HER-Z/neu amplification and
`clinical outcome are in patients with positive axillary lymph
`nodes, we next performed subgroup analyses by lymph node
`status.
`
`Fig. 3. Disease—free survival (DFS) by axillary lymph node status by amplifi-
`cation of HER-Z/neu gene. Node-negative patients with HER-Z/neu amplification
`had marginally better disease-free survival than patients with single copy of the
`gene (P = 0.10). Amplification of HER—Z/neu did not correlate with disease—free
`survival for node-positive patients (P = 0.43).
`946
`
`nAt..uIAAAA.-J :95” nnl-Ln-vann AAnwiA..mAA1A A»... A» ”AAAML‘A» r: 5114‘? 4% 40nd AMAN'AAW AnhAaiailAh {Aw f‘ aaaaa
`
`3
`
`

`

`HER-2/uu AMPLlFlCATlON 1N BREAST CANCER
`
`the literature (Table 1). It might be argued that our choice of
`p53 as a standard of comparison might artificially inflate our
`amplification rate since there is now evidence that alterations
`of the p53 gene can occur in human breast cancer (32, 33). We
`have addressed this issue previously (10) by using separate
`probes for myeloperoxidase and p53. The p53 gene is found on
`the short arm of chromosome 17, while the myeloperoxidase
`gene is found on the long arm of chromosome 17 near the
`HER-Z/neu gene. The use of both probes provided an assess-
`ment of the tumors for duplication of either all or part of
`chromosome 17 and thus addressed any increase in signal
`resulting from duplication of the chromosome. In no case that
`was called amplified did we find evidence for chromosomal
`duplication. Furthermore, if one allele for p53 was deleted, this
`would result in a maximum possibility of overestimating HER-
`2/neu copy number by a factor of 2. However, the presence of
`DNA from normal cells present in every breast tumor specimen
`with normal alleles for p53 would dilute this potential over-
`estimate to a factor of always <2.
`Amplification of the HER—Z/neu gene was independent of
`the axillary lymph node status, the steroid receptor status, and
`the age of the patient but was weakly related to the size of the
`primary tumor. There is little agreement in the literature con-
`cerning the relationships, or lack of relationships, between
`HER-Z/neu amplification and other prognostic factors. Some
`studies have found no association between gene amplification
`and lymph node status (6, 12, 14, 15, 19, 24), while others have
`reported a weak relationship (1 , 5, 7, 21—23, 26). Several studies
`have reported an association with steroid receptor status (7, 13,
`18, 19, 23, 24), but others have failed to confirm this (1, 6, 12,
`22, 26). None of the studies have found relationships with age,
`and only two have reported a correlation with tumor size (23,
`26).
`The strong correlations between HER-Z/neu amplification
`and disease-free and overall survival that were observed among
`our original cohort of 86 node-positive patients could not be
`confirmed in this new group of patients. The actuarial 5-year
`disease—free survival probabilities for the patients in this study
`were 70 i 3% and 65 t 5%, respectively, for patients with
`unamplified and amplified tumors. However, before concluding
`that HER-Z/neu is not a significant prognostic factor for pa-
`tients with primary breast cancer, one must be assured that the
`sample size is sufficient to justify such a conclusion. With 362
`patients at risk for recurrence for at least 5 years, we would
`have approximately 80% power to detect a 15% difference in
`5-year relapse rates between patients with amplified and un-
`amplified tumors [using a two—sided test of proportions at the
`5% level of significance adjusted for a 30% amplification rate
`(31)]. Ideally, a strong prognostic factor should separate pa-
`tients into distinctive categories with different probabilities of
`relapse. While differences <15% may be important from a
`biological point of view, they would only be of marginal utility
`from a clinical point of view.
`The results for the 177 node-negative patients were unambig-
`uously negative. HER-Z/neu amplification was not associated
`with either disease-free or overall survival in either univariate
`
`or multivariate analyses. This finding is in agreement with all
`of the published reports that have attempted to correlate HER—
`2/neu amplification with disease outcome in node—negative
`patients. Although the univariate results for the 185 node-
`positive patients were marginally significant (P = 0.07), the
`significance was not retained in multivariate analyses. This is
`consistent with the report of Borg et a1. (23) in which the strong
`947
`
`correlations observed in univariate analyses disappeared in mul—
`tivariate analyses. Other studies have demonstrated significant
`associations between HER-Z/neu amplification and clinical out-
`come, however, these findings have been confined to node-
`positive patients (1, 3, 14—16).
`The poor evaluability rate of DNA extraction (55%) suggests
`that routinely analyzing nuclear pellets for amplification of the
`HERQ/neu gene will have little clinical utility. One might
`anticipate greatly improved evaluability rates for fresh tumors
`or frozen tumor powders, but the technical aspects of the assay
`may continue to limit its usefulness. However, these limitations
`may not apply to other techniques for measuring oncogene
`abnormalities. For example, several investigations have exam—
`ined the significance of mRNA expression measured by North-
`ern blots (15, 16, 21), and protein expression measured by
`Western blots or immunohistochemical staining (4, 7, 15, 20,
`22, 23, 25, 26, 34—45). Recent reports indicate that, while gene
`amplification is usually accompanied by protein overexpression,
`there is a significant percentage of tumors with a single copy of
`the gene that overexpress the protein product (15, 20, 25, 26,
`41). Thus, measurement of protein expression may provide a
`better indicator of patient outcome.
`A limitation of our study is the uncontrolled administration
`of adjuvant hormonal or chemotherapy to specific patients.
`Even though treatment decisions could not have been based on
`HER-Z/neu amplification status, it is conceivable that treat-
`ment could have confounded the clinical correlations for spe-
`cific patients. Van de Vijver et a1. (35) speculated that over-
`expression of the HER-Z/neu gene might be an early step in
`the development of the tumor rather than an indicator of
`increased metastatic potential. This would be consistent with
`the finding of Heintz et a1. (24) who observed an association
`between HER—Z/neu amplification and increased mitotic activ-
`ity. If overexpression or amplification of the gene is associated
`with an increased growth fraction, then one might hypothesize
`that S-phase-specific chemotherapeutic agents might be partic—
`ularly effective against such tumors. The best way to test this
`hypothesis is to simultaneously measure proliferative rate and
`HER-Z/neu amplification and/or expression in a prospective,
`randomized clinical treatment study. Such studies are now in
`progress in the United States.
`
`REFERENCES
`
`1. Slamon, D. J., Clark, C. M., Wong, S. G., Levin, W. J., Ullrich, A., and
`McGuire, W. L. Human breast cancer: correlation of relapse and survival
`with amplification of the HER—Z/neu oncogene. Science (Washington DC),
`235: 177-182, 1987.
`2. van de Vijver, M., van de Bersselaar, R., Devilee, P., Comelisse, C., Peterse,
`J., and Nusse, R. Amplification of the neu (c-erbBJ) oncogene in human
`mammary tumors is relatively frequent and is often accompanied by ampli-
`fication of the linked c-erbA oncogene. Mol. Cell. Biol., 7: 2019—2023. 1987.
`3. Varley, J. M., Swallow. J. E., Brammer, W. J., Whittaker, J. L., and Walker.
`R. A. Alterations to either c-erbBZ (lien) or c-myc prom—oncogenes in breast
`carcinomas correlate with poor short~tenn prognosis. Oncogene, I: 423—430,
`1987.
`4. Venter, D. J., Tuzi, N. L., Kumar, S., and Gullick, W. J. Overexpression of
`the c~erbB—2 oncoprotein in human breast carcinomas: immunohistological
`assessment correlates with gene amplification. Lancet, 2: 69—71, 1987.
`5. Cline, M. J., Battifora, H., and Yokota, J. Prom-oncogene abnormalities in
`human breast cancen correlations with anatomic features and clinical course
`of disease. J. Clin. Oncol., 5: 999—1006, 1987.
`6. Zhou, D., Battifora, H., Yokota, 1., Yamamoto, T., and Cline. M. J. Asso-
`ciation of multiple copies of the c-erbB—Z oncogene with spread of breast
`cancer. Cancer Res., 47: 6123-6125, 1987.
`7. Berger, M. S., Locher, G. W.. Saurer, S., Gullick, W. J., Waterfield, M. D.,
`Groner, B., and Hynes, N. E. Conelation of c—erbB-2 gene amplification and
`protein expression in human breast carcinoma with nodal status and nuclear
`grading. Cancer Res., 48: 1238—1243, 1988.
`8. Ta], M., Wetzler, M., Joselberg. 2., Deutch, A., Gutman, M., Assaf. D.,
`
`nAlllWlA’h’J/‘A ‘VAM fifififififififififififififil ...... I AAAAAA HAAAML‘Ay-L‘
`
`find—I
`
`(A 4nn4 AMAviAAn AAAAAIanAm law!“ “““““
`
`4
`
`

`

`HER-Z/ntu AMPLIFlCATlON IN BREAST CANCER
`
`Kris, R.. Shiloh, Y., Givol, D., and Schlessinger, J. Sporadic amplification
`of the HERZ/neu protooncogene in adenocarcinomas of various tissues.
`Cancer Res., 48: 1517—1520. 1988.
`Ali, 1. U.. Campbell, G.. Lidereau. R., and Callahan, R. Amplification of c-
`”DB-2 and aggressive human breast tumors? Science (Washington DC), 240:
`1795-1796, 1988.
`Slamon, D. J., and Clark, G. M. Response to technical comment. Science
`(Washington DC), 240: 1796—1798, 1988.
`Fontaine. J., Tesseraux, M., Klein, V., Basten, G., and Blin, N. Gene
`amplification and expression of the neu (c-erbB-Z) sequence in human mam—
`mary carcinoma. Oncology, 45: 360—363, 1988.
`Zhou, D.-J., Ahuja, H., and Cline, M. J. Prom—oncogene abnormalities in
`human breast cancer: c-ERBB-Z amplification does not correlate with recur—
`rence of disease. Oncogene. 4: 105—108, 1989.
`Zeiliinger. R., Kury, E., Czerwenka. K., Kubista. E., Sliutz, G.. Knogler, W.,
`Huber, 1.. Zielinski, C., Reiner, G., Jakesz, R., Stafien. A., Reiner, A., Wrba.
`R., and Spona, J. HER-2 amplification, steroid receptors and epidermal
`growth factor receptor in primary breast cancer. Oncogene, 4: 109—114,
`1989.
`Tsuda, H.,Hirohashi,S., Shimosato,Y., Hirota,T.,Tsugane,S., Yamamoto,
`H.. Miyajima, N., Toyoshima, K., Yamamoto. T., Yokota, J., Yoshida, T.,
`Sukamoto, H., Terada. M., and Sugimura, T. Correlation between long-term
`survival in breast cancer patients and amplification of two putative oncogene-
`coamplification units: h:t-1/i’nt»2 and c-erbB—Z/ear—l. Cancer Res, 49: 3104—
`3108, 1989.
`Slamon, D. J., Godolphin, W., Jones. L. A., Holt, J. A., Wong, S. G.. Keith,
`D. E., Leven, W. J., Stuary, S. G., Udove, J., Ullrich, A., and Press, M. F.
`Studies of the HER-2/Iieu protmncogene in human breast and ovarian
`cancer. Science (Washington DC), 244: 707—712, 1989.
`King. C. R., Swain, S. M., Porter, L., Steinberg. S. M., Lippman, M. E.,
`Gelmann, E. P. Heterogeneous expression of erbB—Z messenger RNA in
`human breast cancer. Cancer Res., 49; 4135.419]. 1989.
`Ro, J., 51.1““an A., Ro, J. y" Blick, M., Frye, 1)., Fraschini, G., pmsche,
`H.. and Hortobagyi.G. c-erbB-Z amplification in node-negative human breast
`cancer. Cancer Res, 49: 6941—6944, 1989.
`Adriane, J., Gaudray, R, Simon, M.-P., Simony-Lafontain, J., Jeanteur, P.,
`and Theillet, C. Proto—oncogene amplification and human breast
`tumor
`phenotype. Oncogene, 4. 1339439; 1989.
`Garcia, 1., Dietrich. P.-Y., Aapro, M., Vauthier, G's Vadas, 1..., and Engel,
`E. Genetic alterations of c-myc, c-erval, and c-Ha»ms protooncogenes and
`clinical associations in human breast carcinomas. Cancer Res, 49: 6675—
`6679, 1989.
`Lacroix, H..1glehart,J. D., Skinner, M. A., and Kraus, M. H. Overexpression
`of erbB-Z or EGF receptor proteins present in early stage mammary carci-
`nmiTaztiissedsflOcriCd 3:"???iggsllglmllgnsgcmd primary tumors and regional
`Tavassoli
`'M gm; i> Familial F. Lock N J Mayne L v and
`K' 1th
`‘ N ”
`bB—i/ " bA
`’
`lift
`t‘
`‘
`. d' “W f l
`’ h" ode
`'r
`am,
`‘ c—er
`cer’
`co-amp ' ica '0“ m m I e ,0
`ymp
`"
`metastasis, and c-myc amplification of high tumour grade in human breast
`“mm” B'- 1- Cm“ 60‘ ”5‘51“ '9”
`,
`Yamada, Y" Yoshimoto, M" Murayama, Y" “89*?” E" M‘m' 5., Ya.
`mamoto, T" Sugano, H..'and.Toyoshima, 1" Association Of elevated expres-
`51011 of the cerbB~2 protein wtth spread of breast cancer. Jpn. J. Cancer Res.,
`80: “92-1198, 1989-
`‘
`Bors- A-- Tandony A- Kn Sissurdssongflu Clark, 0- Mn Fem. M'g Fuquav
`S. A. W., Killander, D., and McGuire, W.
`l. HER-Z/iieu amplification
`predicts poor survival in node—positive breast cancer. Cancer Res., 50: 4332—-
`4337, 1990.
`Heintz, N. H-y Leslie. K- 0-. Rogers, L- Au and Howard, P- L- Amplification
`of the c—erbB-Z oncogene and prognosis of breast adenocarcinoma. Arch.
`PllhOL Lab. Med-. “4: 160-163. I990-
`Hanna, W., Kahn, H. J., Andrulis, 1., and Pawson, T. Distribution and
`patterns of staining of neu oncogene product in benign and malignant breast
`diseases. Modern Pathol., 3: 455-461, 1990.
`Uehara, T., Kanelto, Y., Kanda, N., Yamamoto. T., Higaslii, Y., Nomoto.
`C.. lzumo, T., Takayama, 5., and Sakurai, M. c-erbB-Z and c—erbA—l (ear-1)
`
`10.
`
`ll.
`
`12.
`
`13.
`
`14.
`
`15.
`
`16.
`
`17.
`
`18.
`
`19.
`
`20.
`
`2|.
`
`22.
`
`23.
`
`24.
`
`25.
`
`26.
`
`gene amplification and ccrbB—Z protein expression in Japanese breast can—
`cers: their relationship to the histology and other disease parameters. Jpn. J.
`Cancer Res.. 81: 620-624, 1990.
`27. Clark, G. M., Dressler, L. G., Owens, M. A., Pounds, G., Oldaker. T., and
`McGuire, W. L. Prediction of relapse or survival
`in patients with node—
`negative breast cancer by DNA flow cytometry. N. Engl. J. Med., 320: 627—
`633, 1989.
`28. Kaplan, E. L., and Meier, P. Nonparametric estimation from incomplete
`observations. J. Am. Stat. Assoc., 53: 457—481, 1958.
`29. Mantel, N. Evaluation of survival data and two new rank order statistics
`arising in its consideration. Cancer C