Original article
`
`Annals of Oncology 7:245-250, 1996.
`C1 1996 Kluwn- Aca.dmlic Publiskrs. Primed in tk Netkrlands.
`
`The effect of adjuvant prednisone combined with CMF on patterns of relapse
`and occurrence of second malignancies in patients with breast cancer
`
`G. Marini, S. Murray, A. Goldhirscb, R. D. Gelber, M. Castiglione-Gertsch, K. N. Price,
`M. H. N. Tattersall, C.-M. Rudenstam, J. Collins, J. Lindtner, F. Cavalli, H. Cortes-Funes,
`A. Gudgeon, J. F. Forbes, E. Galligioni, A. S. Coates & H.-J. Senn
`For the Intemarional (Ludwig) Breast Cancu Study Group
`
`Summary
`
`Background: The addition of low-dose prednisone (p) to the
`adjuvant
`regimen of cyclophosphamide, methotrexate,
`5-fluorouracil (CMF) allowed patients to receive a larger
`dose of cytotoxics when compared with those on CMF alone.
`However, disease-free survival and overall survival were
`similar for the two groups. To test the hypothesis that low(cid:173)
`dose prednisone might influence the efficacy of the cytotoxic
`regimen used, the toxicity profiles of the two treatment regi(cid:173)
`mens and the patterns of treatment failure (relapse, second
`malignancy, or death) were examined.
`Patients and methods: 491 premenopausal and perimeno(cid:173)
`pausal patients with one to three positive axillary lymph
`nodes included in International (Ludwig) Breast Cancer
`Study Group (IBCSG) trial I from 1978 to 1981 and ran(cid:173)
`domized to receive CMF or CMFp were analyzed for dif(cid:173)
`ferences in long-term outcome and toxic events. The 250
`patients assigned to CMF and prednisone received on the
`average 12% more cytotoxic drugs than those who received
`CMFalone.
`Results: The 13-year DFS for the CMFp group was 49%
`as compared to 52% for CMF alone, and the respective OS
`percents were 59% and 65%. Several toxic effects such as
`
`leukopenia, alopecia, mucositis and induced amenorrhea
`were reported at a similar incidence in the two treatment
`groups. Using cumulative incidence methodology for com(cid:173)
`peting risks, we detected a statistically significant increase in
`first relapse in the skeleton for the CMFp group at 13 years
`follow-up with a relative risk (RR) of 2.06 195% confidence
`interval (CI), 1.23 to 3.46; P- 0.004). Patients with larger
`tumors in the CMFp regimen were especially subject to this
`increase with a RR for failure in the skeleton of 3.32 (95%
`CI, 1.57 to 7.02; P- 0.0005). CMFp-treated patients also
`had a larger proportion of second malignancies (not breast
`cancer), with RR of 3.34 (95% CI, 0.91 to 12.31; P- 0.09).
`Conclusions: Low-dose continuous prednisone added to
`adjuvant CMF chemotherapy enabled the use of higher
`doses of cytotoxics. This increased dose had no beneficial
`effect on treatment outcome, but was associated with an in(cid:173)
`creased risk for bone relapses and a small, not statistically
`significant increased incidence of second malignancies. The
`effects of steroids, which are widely used as antiemetics (oral
`or pulse injection) together with cytotoxics, should be investi(cid:173)
`gated to identify their influence upon treatment outcome.
`
`Key words: adjuvant therapy, breast cancer, CMF, patterns
`of relapse, prednisone, secondary neoplasm
`
`Introduction
`
`The cyclophosphamide, methotrexate, 5-fluorouracil
`(CMF) regimen has been widely evaluated. Since the
`first report of its success in the adjuvant setting [1] for
`reducing relapses and mortality, the regimen has been
`modified in a variety of ways without sufficient study of
`the influence that these modifications might have on
`treatment outcome. The CMF combination chemo(cid:173)
`therapy was based on Cooper's 1969 communication,
`reporting the effectiveness of the regimen in advanced
`breast cancer (2). This CMF regimen was composed of
`cyclophosphamide (C) given orally for 14 consecutive
`days (100 mg/sqm body surface), with methotrexate
`(M), 40 mg/sqm i.v. followed by 5-fluorouracil (F) 600
`mg/sqm i.v., both on days 1 and 8 of a 28-day course.
`The Cooper regimen used C given daily, continuously,
`and weekly M and F. It also contained weekly vincris-
`
`tine (V) and daily prednisone (P). There have been
`attempts to assess the effectiveness of various combina(cid:173)
`tions of C, M, and F, with and without vincristine and
`prednisone in advanced disease, but the reported
`results by Dr. Cooper have never been replicated (3]. A
`trial in advanced disease suggested that the CMF with
`C given orally every four weeks, as described above,
`provided more effective disease control in terms of
`time to progression and survival when compared with
`the same regimen given intravenously once every three
`weeks [4]. In addition, reducing the dose of CMF was
`shown to be associated with a worse outcome [5).
`In the adjuvant setting, the use of prednisone was
`investigated in a Cancer and Acute Leukemia Group B
`(CALGB) trial of 712 patients who received either
`CMFVP or CMF. The results were only partially re(cid:173)
`ported [6]; premenopausal women with 1-3 N+disease
`had a better disease-free survival and overall survival if
`
`JANSSEN EXHIBIT 2060
`Mylan v. Janssen IPR2016-01332
`
`

`

`
`
`
`
`
`
`

`

`246
`
`they received CMFVP. The Eastern Cooperative On(cid:173)
`cology Group (ECOG) trial of 371 premenopausal
`women [7, 8) showed no statistical difference in out(cid:173)
`come in the group which received CMF with predni(cid:173)
`sone with or without tamoxifen, as compared with
`CMF alone. In this trial prednisone was given in a dose
`of 40 mg/sqm daily for 14 days together with cyclo(cid:173)
`phosphamide.
`The international (Ludwig) Breast Cancer Study
`Group (ffiCSG) conducted a trial (trial I) in 491 pre(cid:173)
`menopausal and perimenopausal women with 1-3 in·
`volved axillary nodes in whom CMF was compared
`with CMF plus low-dose, daily prednisone (CMFp) [9).
`There were two reasons for adding low-dose predni(cid:173)
`sone to the chemotherapy: (1) the empirical perception
`that prednisone improved subjective tolerance, and (2)
`the results of a Canadian trial in which premenopausal
`women who underwent adjuvant oophorectomy had a
`better treatment outcome with the concomitant use of
`prednisone as compared to ovarian ablation alone [10).
`In the first report of the Ludwig study, no significant
`differences in overall survival or disease-free survival
`were detected between the two groups at 4 years' me(cid:173)
`dian follow-up [9). The study did demonstrate that
`higher doses of CMF could be delivered to patients
`who also received low-dose prednisone as compared to
`those who received CMF alone.
`Experiments conducted almost three decades ago
`indicated that steroids may block the enzyme activation
`of cyclophosphamide by inhibiting the microsomal en(cid:173)
`zyme metabolism of the drug (11, 12]. In the current
`evaluation of the trial we wished to determine whether
`there were distinct patterns of metastatic presentation
`in patients who received prednisone compared with
`those who did not. Since the early Ludwig trials were
`conducted during a period when steroids were not
`usually given as antiemetics, a common use of steroids
`today, this trial provides an unconfounded comparison
`of CMFp versus CMF.
`
`Patients and methods
`
`Data from 491 patients with node-positive breast cancer who
`entered the IBCSG (formerly Ludwig) trial I from 1978 to 1981
`were analyzed. All patients had at least a total mastectomy and axil(cid:173)
`lary clearance as the primary treatment Patients did not receive
`adjuvant radiotherapy.
`These pre- and perimenopausal patients with 1 to 3 involved
`axillary lymph nodes were randomly allocated to receive either 12
`cycles of CMF (cyclophosphamide, methotrexate and 5-fluoroura(cid:173)
`cil) or CMFp (CMF with the addition of continuous, low-dose pred·
`nisone) to evaluate the impact of low-dose prednisone added to
`combination chemotherapy. The CMF regimen was composed of
`cyclophosphamide (C) given orally for 14 consecutive days, 100 mgl
`sqm per day; and methotrexate (M), 40 mglsqm i.v. followed by
`5-fluorouracil (F), 600 mglsqm i.v.; both on days 1 and 8 of a 28-
`day course. A continuous administration low-dose prednisone was
`given at a dose of 7.5 mglday (5 mg A.M. and 2.5 mg P.M.). Follow(cid:173)
`up policy was slandardized: clinical, hematological and biochemical
`assessment was required every three months for two years, and
`thereafter every six months until death. Chest X·rays were required
`
`every six months; bone scans were required every six months for two
`years and then annually. All patient data, including all disease and
`survival related events, were reviewed and classified by the medical
`study coordinator (A.G.).
`Disease-free survival was defined as the interval from random(cid:173)
`ization to relapse, the appearance of a second primary cancer (in·
`eluding a contralateral breast cancer), or death, whichever occurred
`first. Evidence for breast cancer events was recorded as acceptable
`in the presence of a positive cytological or histological finding, or of
`tumor progression demonstrated through prospectively defined
`imaging tests. Bone metastases were backdated to when first sus(cid:173)
`pected. Overall survival was defined as time from randomization to
`death from any cause. Survival curves were estimated by the Kaplan(cid:173)
`Meier method (13) and standard errors were calculated using
`Greenwood's formula [14). The log-rank test was used to test for the
`significance of differences between disease-free and overall survival
`curves for CMF ~= CMFp [15J. Gray-Tsiatis linear rank tests
`were used in addition to the log-rank test in order to test for later
`'hazard differences (16). Cumulative incidence functions [17) were
`estimated for each of the competing sites of first failure, and tests of
`differences between treatment groups were conducted (18). All
`;?
`P-values were two-sided. Sites of first relapse were classified accord-
`ing to their impact on prognosis [19): (1) local, regional, and distant §
`soft tissue or nodal metastases; (2) bone alone or with local, re- ~
`gional, or soft tissue or nodal metastases; (3) viscera alone or with g:
`either bone or local, regional or soft-tissue or nodal metastases. ""
`Contralateral breast cancer, second non·breast cancer malignancies, ~
`:::r
`and deaths without malignancies were also recorded as separate
`categories. Any event was considered to be a component of the first -?
`s
`event if diagnosed within a two-month time frame [19).
`=: c
`0 c
`'" C·
`~
`The results of this trial were previously reported in ~
`1985 at 4 years' median follow-up [9]. At that time no ?;
`r2
`significant differences were detected with respect to cr
`disease-free or overall survival. At 13 years' median ~
`follow-up, we are focusing on long term disease-free [4
`and overall survival, and sites of first relapse. Figure 1 g
`displays the Kaplan-Meier plots for disease-free and ~
`overall survival. A non-significant separation in the dis- 3
`ease-free and overall survival probabilities favoring !f
`CMF alone developed after five years' median of fol- J~
`low-up. Table 1 presents the treatment comparisons for ~
`aU patients and the major subgroups of patients includ-
`ing estrogen receptor status, tumor size, and age. The
`13-year disease-free survival (DFS) percentage for the
`
`Results
`
`?(cid:173)
`
`Frgurt! l. Disease-free survival and overall survival according to
`treatment assign;nent for IBCSG (Ludwig) Trial l. The CMF treat(cid:173)
`ment group is indicated by a solid line, and the CMFp group is indi·
`cared by a dotted line. The median follow-up is 13 years.
`
`

`

`
`
`
`
`
`
`

`

`247
`
`Table 1. IBCSG trial I: DFS and OS by treatment within patient
`subpopulations (13 years' median follow-up).
`
`Pa·
`tients
`
`13-Year P.va!ue 1 J..Year P.va!ue
`Failures
`OS
`(deaths) DFS
`percent
`percent
`(%}+s.e.
`(%)+ s.e.
`
`All pat~mt:r
`CMF
`CMFp
`ER-posilive:'
`CMF
`CMFp
`ER·M[laflvt!
`CMF
`CMFp
`ER unknown•
`CMF
`CMFp
`Tumor size <;2 em
`CMF
`CMFp
`Tumor stu:> 2 em
`CMF
`CMFp
`Age< 4()
`CMF
`CMFp
`Age il' 40
`CMF
`CMFp
`
`241
`250
`
`ll4 (83)
`127 (98)
`
`52±3
`49±3
`
`0.39
`
`64
`71
`
`59
`60
`
`H8
`! 19
`
`118
`117
`
`123
`133
`
`55
`61
`
`186
`189
`
`31 (23)
`38 (30)
`
`51± 7
`43± 7
`
`23 (16)
`25 (22)
`
`62±6
`56±7
`
`0.72
`
`0.62
`
`6(}(44)
`64 (46)
`
`47±5
`47±5
`
`0.61
`
`57 (40)
`55 (41)
`
`52±5
`52±5
`
`57 (43)
`72 (57)
`
`52±5
`45±5
`
`0.92
`
`0.25
`
`32 (25}
`38 (25)
`
`41 ± 7
`37±7
`
`0.61
`
`82 (58)
`89 (73)
`
`55±4
`52± 4
`
`0.52
`
`65± 3
`59±3
`
`65±6
`59± 7
`
`73±6
`63±6
`
`61±5
`61±4
`
`65±5
`66± 5
`
`65±4
`53±5
`
`51± 7
`55± 7
`
`68±4
`60±4
`
`0.30
`
`0.50
`
`0.29
`
`0.88
`
`0.92
`
`0.21
`
`0.77
`
`0.17
`
`• ER-posin\'e - > 10 fmoVmg protein, 52% of the patients bad estrogen
`receptors assessed in the trial which recruited patients between 197 8 and
`1981.
`
`CMFp group was 49% compared to 52% for CMF
`alone !relative risk (RR), 1.12, P- 0.39) and the re(cid:173)
`spective overall survival (OS) percentages were 59%
`and 65% (RR, 1.17, P = 0.30). Results were similar for
`the log-rank test and the Gray-Tsiatis test for later
`hazard differences. The log-rank P-values are displayed
`in Table 1. A statistically non-sigrlificant but observable
`decrease in overall and disease-free survival was asso(cid:173)
`ciated with CMFp compared with CMF for patients
`who were older or had larger tumors. This effect was
`not observed in younger patients or patients with
`smaller tumors.
`The proportion of patients who received the indicat(cid:173)
`ed percent of protocol dose during the first 6 courses of
`CMF was larger for those who also received low-dose
`prednisone as clearly indicated in Figure 2. On the
`average, 72% of full CMF dose in cycles 1 to 6 was re(cid:173)
`ceived by CMF-treated patients as compared with 83%
`for the CMFp-treated patients. The incidences of any
`toxicity and severe or worse toxicity were similar in the
`two treatment groups, as shown in Table 2. Fewer pa(cid:173)
`tients who received CMFp experienced leukopenia
`(90% for CMF versus 82% for CMFp; P = 0.03). How(cid:173)
`ever, more patients were recorded to have alopecia on
`the CMFp treatment arm (64% for CMF versus 72%
`for CMFp; P ... 0.04).
`Differences in sites of first relapse, occurrence of
`another primary neoplasia (not breast cancer), and
`death without cancer according to treatment group
`
`% of Full CMF Dose Received
`Cycles 1-6
`Figurt~ 2. Proportion of patients. who received at least the indicated
`average total dose of CMF for each of the two treatment arms. The
`CMF arm is indicated by a solid line, and the CMFp arm is indi(cid:173)
`cated by a dotted line. The areas under the curves represent the
`average amount of CMF received. This figure was reproduced with
`pennission from Cancer Res 1985; 45:4454-9.
`
`Table 2. Incidence of toxicities of any grade and severe or worse
`according to treatment group.
`
`::;->
`0
`3
`
`CMF
`
`CMFp
`
`Any
`Severe
`Severe Any
`grade or
`grade or
`worseb
`worse"
`
`P.value
`any
`grade
`
`Leukopenia
`Nausea/vomiting
`Stomatitis/mucositis
`Diarrhea
`Alopecia
`Induced amenorrhea
`
`90%
`101.
`• ,a
`77%
`14%
`23%
`4%
`20%
`0%
`64% N/N
`84%, N/A'
`
`·82%
`2%
`80%
`13%
`31%
`3%
`1%
`20%
`72% N/A•
`87% N/A•
`
`0.03
`0.77
`0.12
`0.63
`0.04
`0.40
`
`• N/ A, not applicable: severe alopecia and induced amenorrhea
`were not defined.
`b Toxicities of grade 3 or grade 4.
`
`were analyzed (Table 3). Patients who received low(cid:173)
`dose prednisone had a higher percentage of first re(cid:173)
`lapse in the bone and a higher incidence of second
`neoplasia. The incidence of other sites of relapse was
`similar across treatments.
`To further assess this apparent difference between
`the treatment groups with respect to bone recurrences
`and occurrences of second primaries, estimates of the
`cumulative incidence for site of first relapse were cal(cid:173)
`culated for subgroups defined by estrogen receptor sta(cid:173)
`tus, tumor size, and age. There was a statistically signifi(cid:173)
`cant increase in first relapse in the bone alone or bone
`with local, regional, or distant soft tissue/nodal metas(cid:173)
`tases for the CMFp treatment group (RR =- 2.06; 95%
`confidence interval (0)- 1.23 to 3.46; P- 0.004)
`which remained significant after adjusting for multiple
`tests with Bonferroni's method l20j. This increased risk
`of bone relapse was largest among patients with tumor
`size greater than 2 em (RR = 3.32; 95% CI"" 1.57 to
`7.02; P = 0.0005). When first relapse in the bone was
`accompanied by a visceral relapse in the same two
`
`

`

`
`

`

`248
`
`Table 3. lBCSG trial!: Sites of first failure (13 years' median follow
`up) (percentage in parentheses).
`
`Local
`Contra breast
`Regional ± local
`Distant
`Soft tissue/nodes
`Bone
`Viscera<
`2nd primary
`Death w/o recurrence
`Total failed
`Total cases
`
`CMF
`
`CMFp
`
`16 (7)
`12 (5)
`22 (9)
`58 (24)
`2 (1)
`20 (8)
`36 (5)
`3 (t)•
`3 (1)
`l14(47)
`241
`
`15 (6)
`6 (2)
`17 (7)
`77 (31)
`3 (1)
`42 (17)
`32 (13)
`9 (4)0
`3 (1)
`127 (51)
`250
`
`• Ovarian cancer, endometrial cancer, and multiple myeloma (one
`each).
`" Gastric, duodenal, rectal, lung, ovarian, endometrial, bladder,
`thyroid cancers, and Hodgkin's disease (one each).
`" 12 CMF and 10 CMFp categorized as viscera had bone as a con(cid:173)
`comitant recurrence site.
`
`month period no significant treatment differences were
`found. There was a trend for increased incidence of
`
`second primaries in the prednisone treated group (1%
`vs. 4%, RR- 3.34, 95% CI .. 0.91 to 12.31, P- 0.09).
`Comparisons of the cumulative incidence according
`to the treatment group for previously defined sites of
`first relapse are displayed in Figure 3. The cumulative
`incidence of second primaries is also presented. The
`difference in the incidence of bone metastases among
`the two treatment groups appeared early, during the
`second year of follow-up.
`
`Discussion
`
`Low-dose prednisone was added to the combination of
`cyclophosphamide, methotrexate and
`fluorouracil
`(CMF) in order to test its potential beneficial endocrine
`effect. No advantages in disease-free survival (DFS)
`and overall survival (OS) were seen, even though higher
`average doses of CMF were administered when cyto- ~
`toxics and prednisone were given together.
`~
`This finding is not surprising. In randomized studies ~
`of patients with advanced disease, increases in dose- ~
`intensity of 10 to 300% above the 'standard' dose ~
`
`Local, Regional, and/or Distant Soft Tissue or Nodes
`
`Bone
`
`"?
`~-
`§
`0
`;::
`~-
`;;
`~·
`,?
`
`~ = 7 0
`
`~ tr:::
`0"
`'<
`(lq :::
`~
`§
`
`[/J " "0 ;;;
`3 cr
`~
`N
`Y'
`'" 0
`"''
`
`'"\
`
`Fi!JiiTe 3. Sites of first failure: local/regional/distant soft tissue or nodes (top left), bone with or without the previous sites listed (top right),
`visceral with or without the previous sites listed (bottom left), and second primary as the first and only site of first recurrence (bottom right).
`Tne CMF arm is indicated by a solid line, and the CMFp arm is indicated by a dotted line.
`
`

`

`
`

`

`produced, at best, only modest effects on survival [21,
`5]. In the adjuvant setting, a National Surgical Adjuvant
`Breast and Bowel Project (NSABP) trial B-22 inves(cid:173)
`tigating a 100 percent increase in alkylating agent
`intensity and dose on 2238 patients failed to indicate
`that these modifications of an adriamycinlcyclophos(cid:173)
`phamide combination resulted in any disease-free sur(cid:173)
`vival or overall survival advantage within approxi(cid:173)
`mately 3 years of follow-up [22]. Thus, an increased
`cytocidal dose of the order of 5 to 10 fold that can be
`achieved for example with hematopoietic growth fac(cid:173)
`tors and peripheral blood progenitor cell or bone
`marrow support will be required to potentially obtain
`an improvement in results. Furthermore, there is evi(cid:173)
`dence from a Cancer and Acute Leukemia Group B
`(CALGB) trial [231 that, reduction of the 'standard'
`cytotoxic dose is associated with shorter disease-free
`and overall survival.
`Another issue concerning dose and efficacy of cyto(cid:173)
`toxics relates to the possible inhibition of microsomal
`enzyme metabolism of cyclophosphamide. The experi(cid:173)
`mental data suggesting that prednisone might effect
`cyclophosphamide activation (11, 121 were not consid(cid:173)
`ered at the time the study was designed. Leukopenia,
`alopecia, and induced amenorrhea are clinical manifes(cid:173)
`tations mainly due to the effect of activated cyclophos(cid:173)
`phamide. While leukopenia was significantly reduced
`in the prednisone treated group allowing more CMF to
`be given, the occurrence of alopecia and amenorrhea
`did not completely clarify whether the experimental
`fmdings of inhibition of cyclophosphamide metabolism
`by prednisone has clinical relevance.
`A further interesting observation from this study
`was the increased incidence of bone relapses in the
`patients who received CMF with prednisone compared
`with those who received CMF alone. No reports about
`changes in the patterns of relapse in the prednisone
`treated group are available from the Canadian Group's
`trial of oophorectomy followed by prednisone given for
`the duration of five years [10]. The lack of a report on
`increased incidence of bone relapse in the Canadian
`trial might be due to either a failure to investigate this
`aspect, or to a counterbalancing effect of the oophorec(cid:173)
`tomy to decrease tumor progression in the bone. Some
`support for this conjecture comes from the Internation(cid:173)
`al (Ludwig) Breast Cancer Study Group (IBCSG) trial
`IT for premenopausal women with 4 or more axillary
`nodes involved investigating oophorectomy plus CMFp
`compared with CMFp alone. In that trial, the incidence
`of bone metastases observed in the CMFp-treated pa(cid:173)
`tients (especially if their tumors contained estrogen
`receptors) was lower in the group of patients who also
`received oophorectomy (35% versus 42% at 15 years).
`An explanation for increased skeletal involvement
`after a chronic use of low-dose prednisone for one year
`may be related to the effects of this drug upon bone ab(cid:173)
`sorption. Chronic treatment with prednisone is known
`to enhance bone absorption and to interfere with bone
`repair, both resulting in increased osteoporosis. The
`
`249
`
`association between increased bone absorption, osteo(cid:173)
`porosis and frequency of metastases in the bone has
`not been described, but it might be postulated as one
`theoretical mechanism for enhancement of tumor
`growth in the bone (bone-frame hypothesis). The
`mechanism of impaired bone metabolism by glucocor(cid:173)
`ticoids is not fully elucidated, but it is likely that meta(cid:173)
`bolic effects interfere with cytokine-mediated regula(cid:173)
`tion of osteoclast proliferation and function, which may
`play a major role in the process of seeding of metas(cid:173)
`tases.
`The effect of the cytokines on skeletal structure may
`be inhibited by chronic glucocorticoid administration.
`This inhibition of cytokines such as IL-l, IL-6 and TNF
`in the bone might result in the impairment of an organ
`specific natural anti-neoplastic activity which will even(cid:173)
`tually lead to a preferential skeletal metastatic prolif(cid:173)
`eration (cytokine hypothesis).
`Another observation from this trial relates to the in(cid:173)
`creased cumulative incidence of second primary tu(cid:173)
`mors in patients treated with the CMF combination
`and low-dose, continuous prednisone. A possible ex(cid:173)
`planation is that the immunosuppressive effect of pred(cid:173)
`nisone for the duration of one year permits tumor
`proliferation. The duration of immunosuppression,
`however, seems to be too short to exert a direct effect
`on appearance of secondary tumors. The number of
`observed tumors in this study is also too small to recog(cid:173)
`nize if the pattern of neoplasms were typical of those
`observed in the immunocompromised host.
`In conclusion, despite the statistical uncertainty re(cid:173)
`lated to the increased incidence of second neoplasia
`following CMF plus low-dose prednisone, this obser(cid:173)
`vation might have clinical relevance to stimulate assess(cid:173)
`ment of such putative side effects in other series. Spe(cid:173)
`cifically, this should be evaluated where repeated
`pulses of high-dose prednisone are being given as anti(cid:173)
`emetics or as premedication for taxanes with uncertain
`effects upon
`immunosuppression and metabolic
`pathways. The use of continuous, low-dose prednisone
`is not recommended in the adjuvant setting, despite its
`ability to decrease subjective side effects and improve
`tolerance and dose intensity of CMF chemotherapy.
`
`Acknowledgements
`
`We thank the patients, physicians, nurses, and data
`managers, who participate in the International Breast
`Cancer Study Group trials. We also acknowledge Mary
`Isley for her special contributions as Coordinating
`Data Manager for the group. We also gratefully
`acknowledge the initial support provided by the Lud(cid:173)
`wig Institute for Cancer Research and the continuing
`support for central coordination, data management,
`and statistics provided by The Swiss Cancer League,
`The Cancer League of Ticino, The Swedish Cancer
`League, The Australia-New Zealand Breast Cancer
`Trials Group (NH&MRC Grants: 880513, 890028,
`
`N
`0
`0'
`
`

`

`
`

`

`250
`
`910420, 920876), The Australian Cancer Society, The
`Frontier Science and Technology Research Founda(cid:173)
`tion, and The Swiss Group for Clinical Cancer Re(cid:173)
`search (SAKK). This research was partially presented
`at the ESMO meeting in Lisbon, 1994.
`
`References
`
`1. Bonadonna G, Brusamolino E, Valagussa P et al. Combination
`chemotherapy as an adjuvant treatment in operable breast can(cid:173)
`cer. N Engl J Med 1976; 294: 405-10.
`2. Cooper RG. Combination chemotherapy in hormone resistant
`breast cancer. Proc Am Assoc Cancer Res 1969; 10: 15.
`3. Henderson I C. Chemotherapy for metastatic disease. In Harris
`JR, Hellman S, Henderson IC, Kinne DW (eds): Breast Dis(cid:173)
`eases. Second edition. Philadelphia; JB Lippincott Company
`1991; 604-65.
`4. Engelsman E, Klijn JGM, Rubens RD et al. 'Classical' CMF
`versus a 3-weekly intravenous CMF schedule in postmeno(cid:173)
`pausal patients with advanced breast cancer: An EORTC
`Breast Cancer Co-operative Group phase ill trial (10808). Eur
`J Cancer 1991; 27: 966-70.
`S. Tannock IF, Boyd NF, DeBoer Get al. A randomized trial of
`two dose levels of cyclophosphamide, methotrexate, and flu(cid:173)
`orouracil chemotherapy for patients with metastatic breast can(cid:173)
`cer. J Clin Oncol 1 988; 6: 1377-87.
`6. Weiss RB, Korzun AH, Tormey DC et al. Adjuvant chemo(cid:173)
`therapy for breast carcinoma using CMFVP vt!TSUS CMF vt!rsus
`CMF-MER: The nine year results from patients with 1-3 axil(cid:173)
`lary node.s. In Salmon SE (ed): Adjuvant Therapy of Cancer V.
`Orlando: Grune & Stratton 1987; 327-36.
`7. Tormey DC. Clinical results ill: Experience of randomized
`trials without surgical controls. In Senn HJ ( ed): Adjuvant
`Chemotherapy of Breast Cancer. Recent Results in Cancer
`Research. Heidelberg: Springer Verlag 1984; 155-65.
`8. Tormey DC, Gray R, Gilchrist K et al. Adjuvant chemohormo(cid:173)
`nal therapy with cyclophosphamide, methotrexate, 5-fluorou(cid:173)
`racil, and prednisone (CMFP), or CMFP plus tamoxifen com(cid:173)
`pared with CMF for premenopausal breast cancer patients: An
`Eastern Cooperative Oncology Group trial. Cancer 1990; 65:
`200-6.
`9. Ludwig Breast Cancer Srudy Group. A randomized trial of
`adjuvant combination chemotherapy with or without predni(cid:173)
`sone in premenopausal breast cancer patients with metastases
`in one to three axillary lymph nodes. Cancer Res 1985; 45:
`4454-9.
`
`10. Meakin JW, AJlt WEC, Beale FA et al. Ovarian irradiation and
`prednisone following surgery and radiotherapy for carcinoma
`of the breast Can Med AssocJ 1979; 120: 1221-31.
`11. Hayakawa T, Kanai N, Yamada Ret al. Effect of steroid hor(cid:173)
`mone on activation of Endoxan (cyclophosphamide). Biochem
`Pharmacol1969; 18:129-35.
`12. Faber OK, Mouridsen HT. Cyclophosphamide activation and
`corticosteroids. N Engl 1 Med 1974; 291:211.
`13. Kaplan EL, Meier P. Nonparametric estimation from incom(cid:173)
`plete observations. JAm Statist Assoc 1958; 53: 457-81.
`14. Greenwood M. Reports on Public Health and Medical Sub(cid:173)
`jects. Vol. 33. The Narural Duration of Cancer. London: His
`Majesty's Stationery Office 1926; 1-26.
`15. Cox DR Regression models and life tables (with discussion). J
`R Stat Soc B (Methodology) 1972; 34: 187-220.
`16. Gray RJ, Tsiatis AA. A linear rank test for use when the main
`interest is in differences in cure rates. Biometrics 1989; 45:
`899-904.
`17. Kalbfleisch JD, Prentice RL. The Statistical Analysis of Failure
`Time Data. New York: Wiley 1980; 169.
`18. Gray RJ. A class of k-sample tests for comparing the cumu(cid:173)
`;?
`lative incidence of competing risks. Ann Stat 1988; 16: 1141-
`54.
`~
`19. Goldhirseh A, Gelber RD, Castiglione M. Relapse of breast g
`cancer after adjuvant treatment in premenopausal and peri- §:
`menopausal women: Patterns and prognoses. J Clin Oncol
`::;o
`1988; 6:89-97.
`~
`Miller R. Simultaneous Statistical Inference. New York: ~
`*i
`McGraw Hil11966.
`21. Antman K. Dose-intensive therapy in breast cancer. In Armi- 5
`tage JO, Antman KH (eds): High-Dose Cancer Therapy. Balti- §
`more: Williams & Wilkins 1992; 701-18.
`0
`22. Dimitrov N, Anderson S, Fisher B et al. Dose intensification ~
`and increased total dose of adjuvant chemotherapy for breast ~
`cancer: Findings from NSABP B-22. Proc Am Soc Clin Oncol 5
`1994; 13: 58.
`:;
`Wood WC, Budman DR, Korzun AH et al. Dose and dose ;.-.•
`intensity of adjuvant chemotherapy for stage II, node-positive
`breast carcinoma. N Eng! J Med 1994; 330: 1253-9.
`
`20.
`
`23.
`
`Received 12 December 1995; accepted 6 February 1996.
`
`0 "
`
`Correspondence to:
`A ron Goldhirsch, M.D.
`International Breast Cancer Study Group
`Ospedale Civico
`6900Lugano
`Switzerland
`
`

`

`
`