High dose chemotherapy and autologous stem cell transplantation as adjuvant therapy for primary breast cancer patients with four or more lymph nodes involved: long-term results of an international randomised trial

R. C. Coombes1,*, A. Howell3, M. Emson4, C. Peckitt2, C. Gallagher5, C. Bengala6, A. Tres7, R. Welch3, P. Lawton8, R. Rubens9, E. Woods4, J. Haviland2, D. Vigushin1, E. Kanfer1, J. M. Bliss2 On behalf of the International Collaborative Cancer Group (ICCG)

1 Cancer Research UK Department of Cancer Medicine, Imperial College London, Hammersmith Hospital, Du Cane Road, London; 2 Clinical Trials and Statistics Unit (ICR-CTSU), Institute of Cancer Research, Sutton, Surrey; 3 Cancer Research UK Department of Medical Oncology, Christie Hospital NHS Trust, Wilmslow Road, Manchester; 4 International Collaborative Cancer Group, Department of Medical Oncology, Division of Medicine, Imperial College London, Charing Cross Campus, Fulham Palace Road, London; 5 Department of Medical Oncology, St Bartholomew's Hospital, 1st Floor KGV Building, London UK; 6 University Hospital, Department of Oncology and Haematology, Division of Medical Oncology, Modena, Italy; 7 Hospital Clinico Universitario Lozano Blesa, Servicio de Oncologia Médica, Zaragoza, Spain; 8 Mount Vernon Hospital Cancer Centre, Rickmansworth Road, Northwood; 9 Department of Clinical Oncology, Guy's Hospital, London, UK

* Correspondence to: R. C. Coombes, CR UK, Department of Cancer Medicine, Imperial College London, Hammersmith Hospital, 6th Floor, MRC Cyclotron Building, Du Cane Road, London W12 0NN, UK. Tel: +44-20-8383-5828; Fax: +44-20-8383-5830; Email: c.coombes{at}imperial.ac.uk


    Abstract
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Background:: The purpose of this study was to assess whether a short course of anthracycline containing chemotherapy followed by high dose therapy with autologous stem-cell support improves disease-free and overall survival as compared with conventional, anthracycline containing chemotherapy, in patients with primary breast cancer and four or more histologically involved lymph nodes.

Patients and methods:: Two hundred and eighty one patients entered into a randomised clinical trial were allocated to receive standard, conventional treatment (5-fluorouracil, epirubicin and cyclophosphamide–FEC for six cycles) or FEC for three cycles followed by high dose therapy consisting of cyclophosphamide, thiotepa and carboplatin and stem cell rescue (HDT). To be eligible, patients had to be free of overt metastatic disease and be ≤60 years of age. Analyses were according to intention to treat.

Results:: At a median follow up of 68 months, 118 patients have experienced a relapse or death from breast cancer (62 in the FEC followed by HDT arm and 56 in the conventional FEC arm) and a total of 100 patients have died (54 in the FEC followed by HDT arm and 46 in the conventional FEC arm). No significant difference was observed in relapse-free survival [hazard ratio 1.06, 95% CI 0.74–1.52, p = 0.76] or overall survival [hazard ratio 1.18, 95% CI 0.80–1.75, p = 0.40]. Five patients died from treatment related causes, three as a consequence of HDT and two in the conventional FEC arm.

Conclusions:: At the present time, no benefit has been observed from replacing three cycles of conventional chemotherapy with the HDT regimen described here. Patients should continue to receive conventional chemotherapy as adjuvant therapy for breast cancer.

Key words: autologous stem cell transplantation, high dose chemotherapy, primary breast cancer


    Introduction
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Over the past 20 years, several studies of high dose chemotherapy with haemopoietic-progenitor-cell support have indicated a role for this form of treatment in metastatic breast cancer. Our own study [1Go] and those of others [2Go–6Go], have all indicated that patients with metastatic breast cancer respond to high dose treatment, although only a subset of trials indicated any survival advantage and with a single exception [2Go], these studies were not randomised.

In 1993 Peters et al. [7Go] reported that a non-randomised prospective series of patients with primary breast cancer who had 10 or more tumour positive axillary lymph nodes and who had received high dose therapy with cisplatin, cyclophosphamide and carmustine, had a substantially superior event-free survival than would be expected by comparison with a historical control group. With a median follow-up of 2.5 years, 72% (95% confidence interval, CI 56% to 82%) were event-free in the prospective series compared with historic series of similar patients whose corresponding event-free survival rates ranged from 38% to 52%. This provocative study highlighted the need for randomised trials to confirm or refute these potentially exciting results.

The International Collaborative Cancer Group (ICCG), along with other trials groups subsequently designed randomised trials in patients with poor risk primary breast cancer. By the end of 2002, three of these studies had reported efficacy results [8Go–10Go]. None of these studies revealed any benefit overall for high dose therapy in comparison with a control group. The high dose chemotherapy regimens used by two studies [9Go, 10Go] consisted of high dose cyclophosphamide, thiotepa, and carboplatin, whereas a further study [8Go] used high dose cyclophosphamide, etoposide and cisplatin.

The ICCG HDT trial began in 1993, and closed to recruitment in July 2001 by which time, results from other studies were being made available and the resulting interest in evaluation of high dose therapy in poor-risk primary breast cancer had diminished.


    Patients and methods
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Study design
The ICCG HDT trial was a phase III randomised controlled trial comparing short duration conventional anthracycline containing chemotherapy followed by high dose chemotherapy with a standard course of conventional anthracycine containing chemotherapy. The control arm was a 5-fluorouracil, epirubicin and cyclophosphamide (FEC) regimen that has been described previously as being superior to intravenous CMF [11Go]. The experimental arm consisted of three cycles of the same FEC chemotherapy as used in the control arm followed by the high dose cyclophosphamide, thiotepa and carboplatin (i.e. CTCb) regimen with peripheral blood stem cell support (HDT). The study, therefore, was designed to determine the benefit of substituting a high dose therapy regime for cycles 4–6 of FEC in the adjuvant setting. The primary end points were relapse-free survival, event-free survival and overall survival.

The secondary end points were toxicity experienced during and after treatment. All toxicities were graded on day 1 of each course of chemotherapy according to the National Cancer Institute Canada Common Toxicity Criteria (Version 2). Toxicities specifically assessed included nausea, vomiting, stomatitis, infection and diarrhoea. Other adverse events were also graded.

Eligibility
Patients were eligible if they were ≤60 years of age and had histologically confirmed breast cancer with four or more involved nodes in the axilla. Patients had to have had complete surgical resection of the tumour and axillary node clearance and be free of distant metastases as judged by bone scanning, liver ultrasound and chest X-ray. Patients with abnormal bone marrow, renal or hepatic function (defined as having a white cell count of <4000; platelets <1 000 000, a creatinine clearance of at least 60 ml/min and liver function tests less than 1.5 times the upper limit of normal) or with a WHO performance status of more than 1 were excluded.

Written informed consent was obtained from all patients for enrolment according to institutional guidelines, after having received comprehensive information (both verbal and written). The protocol was approved by the local research ethics committee for each participating hospital.

Randomisation procedure
Centres randomised their patients by telephoning the ICCG Data Centre. The randomisation method used was adapted minimisation, where the weighted probabilities ensure a random component to the allocation. Stratification factors were centre, menopausal status and number of axillary nodes involved (4–9, 10+).

Treatment schedule
The control arm (FEC) consisted of 5-fluorouracil 600 mg/m2 day 1, epirubicin 50 mg/m2 on day 1 and cyclophosphamide 600 mg/m2 day 1 on cycle 1, repeated 3 weeks later on a day 1 and 8 (omitting epirubicin on day 8) basis, for 5 consecutive monthly cycles. The experimental arm consisted of 5-fluorouracil 600 mg/m2 day 1, epirubicin 50 mg/m2 on day 1 and cyclophosphamide 600 mg/m2 day 1 (same as control group), followed by the peripheral blood stem cell harvest at day 11 preceded by 5 µg/kg GCSF on days 2–11. This was then followed by cycles 2 and 3 as for the control arm followed by a continuous infusion over 4 days of cyclophosphamide 6 g/m2, thiotepa 500 mg/m2 and carboplatin 800 mg/m2 (HDT). Autologous stem cells were re-infused 7 days after the start of the conditioning therapy, and patients were nursed in a single room during the period of neutropenia.

Initially, before the results of the earlier ICCG trial were known [11Go], the intent had been for the conventional regimen to consist of treatment with epirubicin (75 mg/m2) and cyclophosphamide (750 mg/m2), omitting the 5-fluourouracil. Nineteen patients (nine control and 10 experimental arms) were treated according to this protocol, prior to the subsequent protocol amendment. The high dose component of the treatment was not affected by this amendment.

Peripheral blood stem cell harvesting was performed by standard methods using a continuous flow cell separator programme with mononuclear separation protocol. The minimum harvested cell counts required to proceed to high dose therapy were a mononuclear cell count of 2 x108 per kg and CD34 cell count of 3 x 106 per kg. If no recovery had occurred on day 14, GCSF was given at 10 µg/kg/day by daily injection. MESNA was given at dose of 2 g/m2 by continuous infusion prior to each dose of chemotherapy and for a further six doses, 2 days after the end of the chemotherapy infusion.

In the high dose therapy arm, anti-emetics were used prophylactically and included intravenous granisetron as well as dexamethasone 8 mg i.v. administered twice daily. After 40 patients had been entered into the trial, and early toxicity observed, the protocol was amended to include routine use of dexamethasone to prevent hepatic veno-occlusive disease, as reported by others [12Go]. All patients received prophylactic antibiotics consisting of ciprofloxacin and fluconazole, which were administered until neutrophil engraftment (>0.5 x109/l). All blood and blood products were irradiated to at least 2500 cGy to prevent graft-versus-host disease.

After chemotherapy, patients were considered for radiotherapy treatment according to protocols determined by each institution. For patients who had been treated by breast conservation surgery, 50 Gy was delivered to the breast in 25 fractions, with 10 Gy boost to the tumour bed. The same dose was given to the cervico-axillary chain. Patients who had been treated by mastectomy and yet required radiotherapy received 50 Gy in 25 fractions to the chest wall and supraclavicular fossa. Tamoxifen 20 mg daily was planned for a total of 5 years for all patients, irrespective of the oestrogen receptor status, commencing after chemotherapy.

Following the completion of chemotherapy treatment, patients were seen 3 monthly until 2 years after randomisation, 6 monthly until 5 years and annually thereafter. No imaging was done on follow-up, reflecting the practice at that time in participating centres.

Statistical methods
All analyses were by intention to treat; patients with protocol deviations were not excluded from the analysis. Interim analyses and each treatment-related death were supplied periodically, and in confidence, to the independent data monitoring committee by the trial statisticians.

Relapse-free survival was measured from the date of randomisation to the date of first recorded local or distant recurrence, contralateral breast cancer or death from breast cancer (without a previously reported relapse). Event-free survival was measured from the date of randomisation to the date of first recorded local or distant recurrence, contralateral breast cancer or death from any cause; thus incorporating deaths from other causes, such as those related to treatment in the absence of a breast cancer relapse. Overall survival was likewise measured from the date of randomisation.

Survival curves were constructed by the Kaplan–Meier method and compared by the log rank test. Cox regression methods were used to investigate whether any clinical prognostic factor was confounding the treatment effect [13Go]. All P values were two-sided. Hazard ratios with values > 1.0 were indicative of a disadvantage to the experimental, high dose therapy, arm.

Toxicity data, presented as simple proportions of subjects experiencing each toxicity, were analysed separately in the first phase (cycles 1–3) and second phase (cycles 4–6) of the trial. Chi-squared tests were used to compare the percentage of grade 3 or 4 toxicities between randomised groups.

With 90% power ({alpha}=0.05), approximately 1000 patients would have been required to detect an improvement in 5-year survival from 30% to 40%. Due to the foreseen difficulties for a small collaborative group of recruiting such a large number of patients in a reasonable time frame, it was decided a priori to recruit 300 patients into this study, with a stated intention for a prospective meta-analysis with a parallel UK trial. Three hundred patients would be sufficient to detect an improvement from 30% to 45% with 80% power ({alpha}=0.05). The study was eventually closed to entry after 281 patients had been recruited.


    Results
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Patients
The ICCG HDT trial commenced accrual in November 1993 and the trial closed to recruitment in July 2001. Patient entry was terminated when the accrual rate dropped off substantially following the disappointing early reports from companion trials evaluating high dose therapy in primary breast cancer [8Go–10Go]. Two hundred and eighty-one patients were randomised into the trial, from eight European hospitals. One hundred and thirty-eight were randomised to receive conventional FEC and 143 to receive a short course of FEC followed by HDT. Two patients were lost to follow-up following randomisation. The median follow-up period for all subjects at the time of this analysis was 5.82 years (IQR: 4.28–7.11 years) and we had received information on all but 35 surviving patients in the last 2 years. Of the randomised subjects, 206 (73%) were followed-up for 36 months and 158 (56%) were followed-up for 48 months. The profile of the trial is shown in Figure 1 (as a consort diagram), indicating the progress of patients through the course of the trial.



View larger version (30K):
[in this window]
[in a new window]
 
Figure 1. Flow of patients through the trial. Of 281 patients, only two were excluded from the final analysis; 268 patients received the allocated treatment.

 
Patient characteristics
Table 1 describes the patient characteristics. The trial cohort represents a relatively young group of patients compared with the typical profile of breast cancer incident cases; the median age was 48 years (range 24–60) in the conventional FEC arm and 46 years (range 27–60) in the short course FEC followed by HDT arm. One hundred and ninety-eight patients (70%) were premenopausal at entry into the trial. The treatment groups were well balanced, at baseline, for clinical characteristics. Thus, 64 (46%) patients in the conventional FEC and 65 (45%) patients in the short course FEC followed by HDT, respectively, had 10 or more lymph nodes involved. The majority of patients had tumour size between 2 and 5 cm (56% of those with known tumour size); however, 34 patients (13% of those with known tumour size) had tumours greater than 5 cm in diameter. Three patients, one in the FEC arm and two in the FEC followed by HDT arm, were subsequently found to have had metastatic disease at the time of randomisation.


View this table:
[in this window]
[in a new window]
 
Table 1. Demographic and clinical characteristics at randomisation

 
Of the 175 patients who had a wide local excision, 156 patients (89%) also received radiotherapy to the breast. In addition, of those who had a mastectomy (96 patients), 70 received chest wall radiation [38 patients (79%) in the conventional FEC arm; 32 (66%) in the FEC followed by HDT arm]. Radiation to the axilla was given in 26 (19%) of those allocated conventional FEC, compared with 27 (20%) of those allocated to FEC followed by HDT. Tamoxifen was given to 240 patients (86%) in the study, equally balanced in both groups [117 patients (85%) in the conventional FEC arm; 123 (87%) in the FEC followed by HDT arm].

Treatment compliance
Of 143 patients assigned to FEC followed by HDT, 10 refused and one patient received six cycles of FEC in error. Two developed metastatic disease during the course of FEC induction chemotherapy, two patients had emergence of recurrent disease and three patients died as a result of the HDT treatment schedule. Six patients experienced toxicity, which led to early discontinuation of treatment. In the conventional FEC group, 138 were assigned FEC but two subsequently refused treatment. Three patients developed recurrent disease during the chemotherapy, three were deemed too ill (not cancer related) to continue and two died during the FEC treatment. Fourteen patients experienced toxicity, which led to early discontinuation of treatment. Finally 11 patients (five FEC followed by HDT and six conventional FEC) discontinued their chemotherapy early for other reasons. Therefore, 219 completed their allocated course of treatment (Figure 1); 105 in the conventional FEC arm and 114 in the short course FEC followed by HDT arm of the trial.

Relapse-free and overall survival
Of the 279 patients with follow-up, 118 have had a ‘first event’. Eight patients experienced an isolated local relapse (with no distant relapse yet reported), 108 patients had isolated distant metastases (including preceding or concurrent local relapse in 16 patients), one patient had disease in the contralateral breast, and a further patient died of breast cancer without a date of first recurrence being identified. Fifty-six of these patients were allocated conventional FEC, compared with 62 patients allocated to receive short course FEC followed by HDT. The hazard ratio for relapse-free survival was 1.06 (95% confidence interval, CI 0.74–1.52, P=0.76; Figure 2). The 5-year relapse free survival was 59% (95% CI 50% to 67%) and 57% (95% CI 48% to 65%) in the conventional FEC and FEC followed by HDT groups, respectively. Including deaths from any cause prior to relapse (i.e. treatment related or intercurrent deaths) in an event-free survival analysis gave a hazard ratio of 1.07 (95% CI 0.75–1.52), with 59 patients having an event in the conventional FEC arm compared with 66 patients in the FEC followed by HDT arm (P=0.70).



View larger version (22K):
[in this window]
[in a new window]
 
Figure 2. Kaplan–Meier relapse-free survival curves for patients in the short course FEC followed by high dose therapy and conventional FEC arms of the study.

 
Two patients developed non-breast second primary cancers; both were ovarian carcinomas (one in the conventional FEC arm and one in the FEC followed by HDT arm). Only the ovarian carcinoma in the conventional FEC arm occurred in the absence of either local or distance recurrence. No cases of acute myeloid leukaemia or other haematological malignancies were observed.

One hundred patients died (Table 2b): 46 patients allocated to conventional FEC compared with 54 in the group allocated FEC followed by HDT. Of those who died, 92 patients died of breast cancer (43 patients allocated to conventional FEC and 49 allocated to FEC followed by HDT). The other eight died of other causes (three allocated to conventional FEC and five allocated to FEC followed by HDT), without a prior breast cancer relapse. Five of these patients died of treatment-related causes. Three patients in the FEC followed by HDT arm were judged to have died from side-effects of treatment (after completing the allocated treatment). Two of these developed hepatic veno-occlusive disease and died from liver failure, 4 and 5 months after randomisation, and one patient died from cardiomyopathy 4 months after randomisation. In addition, one patient in the FEC arm died from neutropenia associated with colitis 3 months after randomisation; another FEC patient died from acute renal failure 15 months post-randomisation. A further FEC patient died suddenly 10 days after the fourth course of chemotherapy, possibly due to anthracycline-induced arrhythmia. No abnormality was found at autopsy. Apart from these treatment-related deaths, all other deaths from ‘other’ causes, in the absence of a breast cancer relapse, occurred more than 1 year after randomisation.


View this table:
[in this window]
[in a new window]
 
Table 2. Recurrences and deaths by randomisation arm

 
The hazard ratio for overall survival was 1.18 (95% CI 0.80–1.75, P=0.40; Figure 3). The 5-year overall survival was 67% (95% CI 58% to 74%) and 66% (95% CI 57% to 74%) in conventional FEC and FEC followed by HDT groups, respectively. Sensitivity analyses, excluding the initial 19 patients who were randomised to receive chemotherapy without 5-fluorouracil produced results similar to the intention-to-treat analysis.



View larger version (21K):
[in this window]
[in a new window]
 
Figure 3. Kaplan–Meier overall survival curves for patients in the short course FEC followed by high dose therapy and conventional FEC arms of the study.

 
Figures 4 and 5 show relapse-free and overall survival for patients with 4–9 and >10 involved nodes. No heterogeneity was observed between nodal groups.



View larger version (15K):
[in this window]
[in a new window]
 
Figure 4. Forrest plots showing hazard for relapse-free survival in patients with 4–9 involved nodes and those with 10+ nodes involved.

 


View larger version (16K):
[in this window]
[in a new window]
 
Figure 5. Forrest plots showing hazard ratios for overall survival in patients with 4–9 involved nodes and those with 10+ nodes involved.

 
Toxicity
Toxic side-effects associated with conventional FEC in the induction phase of the treatment (cycles 1–3) are well known and were similar in both arms of the study. Grade 3 and 4 nausea and vomiting (15% FEC, 9% HDT), hair loss (6% FEC, 7% HDT) and leucopenia (9% FEC, 15% HDT) were observed in both groups of patients in the induction phase. During this phase, a single patient in each group suffered a deep venous thrombosis and two patients in the group scheduled to subsequently receive HDT arm developed a pulmonary embolism.

Analysis of the second phase (i.e. cycles 4–6 of FEC treatment in the conventional FEC arm and the HDT phase of the FEC followed by HDT arm) showed, as expected, in the 248 patients with complete toxicity data, severe side-effects (grade 3 and 4) in the patients who received HDT including nausea and vomiting (46%), leucopenia (100%), thrombocytopenia (100%) and 24% developed neutropenic fever. Other grade 3 and 4 side-effects were seen in 28% of HDT patients compared with only 5% of patients continuing the conventional FEC regimen (Table 3). Serious side-effects observed during or shortly after HDT included three patients with severe cardiac arrhythmias, two patients with thromboembolic episodes (one pulmonary embolus and one deep venous thrombosis) and one patient developed respiratory failure requiring ventilation. Patients in the FEC arm had similar side-effects in the second phase as in the first (induction) phase of treatment. The patient in the conventional FEC arm who had a deep venous thrombosis in the induction phase had a second deep venous thrombosis when she developed recurrent disease in the second phase of the FEC treatment. Deaths due to side-effects have been described above.


View this table:
[in this window]
[in a new window]
 
Table 3. Toxicity during the second phase, cycles 4–6

 
Toxicities following completion of chemotherapy were also monitored, with 62 patients from the conventional FEC group and 73 from the group allocated to receive FEC followed by HDT experiencing one or more late toxicities. In 73 instances these were considered to be due to tamoxifen (34 in conventional FEC arm and 39 in FEC followed by HDT arm). However, there were 210 (96 conventional FEC and 114 FEC followed by HDT) other long-term toxic events recorded in these patients. The most notable of which were: seven patients suffering thromboembolic events (six conventional FEC, one FEC followed by HDT), eight patients experiencing cardiac events (five conventional FEC, three FEC followed by HDT) and 44 patients reporting musculo-skeletal (principally anthralgia) problems (25 conventional FEC and 19 FEC followed by HDT). Other long-term side-effects included evidence of persisting oedema (eight conventional FEC, 14 FEC followed by HDT) and various infective episodes (six conventional FEC, 22 FEC followed by HDT).

Eight of the above-described late toxicities were thought to be life-threatening events. Two patients, who were randomised to receive conventional FEC, developed pulmonary embolism and six patients, who received FEC followed by HDT, developed persistent pancytopenia (2); pneumonia (1); liver failure due to veno-occlusive disease (1); acute haematemesis (1); and acute renal failure (1).

All pre-menopausal patients who received high dose therapy developed amenorrhoea following completion of chemotherapy.

Harvest and recovery of bone marrow function in the high dose therapy group
Complete data on recovery of blood counts in the patients who received stem cell rescue were available for 118 out of 132 patients who received short course FEC followed by HDT. The median time to recovery (neutrophil count >0.5 x 109) was 10 days with a range of 0–43 days (interquartile range, IQR 9–12). For platelets, the median days to recovery (>20 x 109) was 9 days with a range of 0–52 days (IQR 7–12).


    Discussion
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
The principal finding of this study is that, following three cycles of FEC chemotherapy, high dose chemotherapy with CTCb and peripheral stem cell support does not appear to improve disease-free or overall survival when compared with continuing conventional-dose chemotherapy with a further three cycles of FEC. The CTCb regime is associated with an increase in Grade 3 or 4 toxicities of various types, but despite this, long-term toxicity was similar and there were similar numbers of treatment-related deaths in both arms of the study. This treatment-related mortality rate is similar to other studies using CTCb [10Go, 14Go], and is lower than that reported in earlier high dose chemotherapy studies. Treatment-related deaths with conventional anthracycline chemotherapy are rare but not unknown [15Go]. Although compensated to some degree by the long follow-up, our study suffered the drawback of having smaller numbers of patients than some studies, including the Dutch Adjuvant Study [14Go] in which 885 patients were randomised, the CALGB study [16Go] in which 783 patients were randomised, and the Scandinavian study [10Go] of 525 patients. None of these studies have, however, demonstrated a clear benefit for high dose therapy in primary breast cancer. More recently, the Anglo–Celtic trial of 605 patients has been published [17Go] and this also failed to show any significant benefit for high dose therapy using high dose cyclophosphamide. Other studies published only in abstract form include: Pegase 01 trial, the French study, using cyclophosphamide, mitozantrone and melphelan, where the 3-year disease-free survival for the HDT arm was 71% compared with 55% (P<0.003) [18Go]; and the Milan study, which enrolled 398 patients with similar 5-year disease-free survival and overall survival for the HDT and conventional arms (65% and 62%; 76% and 77%, respectively) [19Go].

More recently the CMA/Pegase 04 trial [20Go] has been published: in this study, although of only 61 patients, an improvement in overall survival was observed.

It is possible that earlier studies [10Go] failed to show a benefit from HDT because patients in the control arm received higher doses of FEC chemotherapy than in the current study, thus potentially masking any benefit from HDT. It is also possible that the reason for a negative result found by the Anglo–Celtic group [17Go] was because high dose cyclophosphamide alone was used.

The ICCG HDT trial study was closed to entry after only 281 patients had been recruited. It is quite possible that a clinically important difference exists which is undetectable with any reliability in a trial of this size. Retrospectively, it has been calculated that the study had 74% power to detect a 15% improvement in 5-year relapse-free survival from 58% to 73%, and 41% power to detect a 10% improvement in relapse-free survival from 58% to 68%.

One further possible reason for failure to show benefit of HDT is that the peripheral stem cells that we infused could have potentially contained micrometastases. However, a study in which no peripheral stem cell rescue were used, that is, where patients were given high dose therapy but without progenitor cell support, also failed to confirm any benefit for high dose therapy [21Go].

The ICCG HDT trial was designed at a time when uncontrolled data from several centres indicated that high dose therapy would be associated with a substantial improvement in survival. At the time of commencing the trial, we had just published a study suggesting that 5-fluorouracil, epirubicin, cyclophosphamide regimen was superior to intravenous cyclophosphamide, methotrexate and 5-fluorouracil [11Go]. We therefore used this regime as the control arm in this study. Like the Swedish group [10Go], the ICCG also considered that it was not appropriate, on the basis of information available at that time, to subject patients to chemotherapy regimens that lasted longer than 6 months. The ICCG HDT trial, therefore, was intended to determine whether three cycles of conventional anthracycline containing chemotherapy (i.e. 3 months) could be compensated for by the addition of a single course of high dose chemotherapy. Consequently, the trial assessed not whether the addition of HDT improved outcome, but rather whether the substitution of HDT for the final three cycles of FEC improved outcome. As a result of this, the total dose of anthracycline in the HDT arm of the study is less than the majority of other studies.

It is difficult to compare the various trials assessing high dose therapy in poor risk primary breast cancer because of the heterogeneity of the published high dose regimens. The regime used in the ICCG HDT trial is similar to the Swedish study [10Go], which showed that patients who received a similar CTCb regimen had a worse disease-free survival when compared with those who received a ‘tailored’ FEC regimen with G-CSF support. The Anglo–Celtic group used high dose cyclophosphamide, the CALGB used combination-alkylating agents. Several studies have used CTCb [8Go, 10Go, 13Go] in a regimen similar to the one used here. As discussed above, no trial has as yet reported a clear benefit for high dose therapy, although longer follow-up may reveal clinically important improvements. If high dose therapy does offer such benefits, then it is likely that a meta-analysis of the trials will be informative. In such a scenario, variations between high dose regimens and indeed between individual conventional regimens must be smaller than the gross difference between conventional and high dose chemotherapy. Indeed, a Cochrane Review has been carried out to include nine studies, entering 3543 patients into trials of high dose therapy in primary breast cancer [22Go]. Although currently unable to use individual patient data, this systematic review is proving informative and work is ongoing. To date, the review reports treatment mortality rates: there were 48 non-cancer deaths in the HDT arm compared with only four in the conventional arm; however, 32 of these deaths were in the CALGB study.

The types of side-effects observed in the ICCG HDT trial have all been reported previously in other studies of high dose chemotherapy. Hepatic veno-occlusive disease was only seen in the earlier phase of the study. After the protocol was changed to include steroids in the chemotherapy regimens, no further episodes of this complication were observed. The beneficial effect of steroids has been noted by others [12Go]. We also saw several other severe side-effects such as thromboembolism and severe pulmonary infection. One patient had to be ventilated as a result of respiratory failure due to interstitial lung disease.

Quality of life was not documented in our study, but others, including the CALGB [16Go] and the Dutch Intergroup study [14Go], have demonstrated significantly worse quality of life for patients receiving high dose therapy at 3–6 months, but by 12 months these differences had narrowed and were no longer significantly different.

In the future, through the use of techniques such as gene expression profiling, it may be possible to identify patients who may benefit from intensive treatment. Indeed, the recently reported results of the Dutch trial [14Go] suggest that a subset of patients may benefit from such treatment. The report suggests that patients with 10 or more positive nodes, younger patients, those with c-erbB2 negative tumours, and those who had a low grade tumour, showed a beneficial effect of high dose therapy on relapse-free survival. A further recently published study, which documented the use of a high dose therapy regimen using cyclophosphamide and thiotepa [23Go], indicated that whilst there was no superiority of the HDT regimen overall, the patients who fulfilled strict entry criteria did appear to have a longer time to recurrence.

It has been estimated that more than 40 000 women with primary breast cancer received high dose therapy outside randomised trials in the 1990s, at a time when no randomised trials were available [21Go]. Now, as information from randomised controlled trials has emerged, each has failed to demonstrate any benefit for such a treatment strategy. This emphasises the need for high-level evidence before new treatments, however promising they seem, are adopted into routine clinical practice. At the present time, we cannot recommend the use of high dose therapy as adjuvant therapy for breast cancer.


    Acknowledgements
 
RCC and JMB received research funding from Cancer Research UK (CR UK). Pfizer and Amgen provided research support for the study. We are grateful to the many research staff who have worked within the participating hospitals and the ICCG Data Centre to achieve a successful trial. Most importantly, we are indebted to the patients who kindly agreed to participate.

Received for publication September 16, 2004. Revision received January 25, 2005. Accepted for publication February 1, 2005.


    References
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
1. Vincent MD, Powles TJ, Coombes RC et al. Late intensification with high-dose melphalan and autologous bone marrow support in breast cancer patients responding to conventional chemotherapy. Cancer Chemother Pharmacol 1988; 21: 255–260.[ISI][Medline]

2. Stadtmauer EA, O'Neill A, Goldstein LJ et al. Conventional dose chemotherapy compared with high-dose chemotherapy plus autologous hematopoietic stem-cell transplantation for metastatic breast cancer. Philadelphia Bone Marrow Transplant Group. N Engl J Med 2000; 354: 1069–1076.

3. Peters W, Shpall E, Jones R et al. High-dose combination alkylating agents with bone marrow support as initial treatment for metastatic breast cancer. J Clin Oncol 1988; 6: 1368–1376.[Abstract]

4. Wallerstein RJ, Spitzer G, Dunphy F et al. A phase II study of mitoxantrone, etoposide, and thiotepa with autologous marrow support for patients with relapsed breast cancer. J Clin Oncol 1990; 8: 1782–1788.[Abstract]

5. Dunphy FR, Spitzer G, Buzdar AU et al. Treatment of estrogen receptor-negative or hormonally refractory breast cancer with double high-dose chemotherapy intensification and bone marrow support. J Clin Oncol 1990; 8: 1207–1216.[Abstract]

6. Bowers C, Adkins D, Dunphy F et al. Dose escalation of mitoxantrone given with thiotepa and autologous bone marrow transplantation for metastatic breast cancer. Bone Marrow Transplant 1993; 12: 525–530.[ISI][Medline]

7. Peters W, Ross M, Vredenburgh J et al. High-dose chemotherapy and autologous bone marrow support as consolidation after standard-dose adjuvant therapy for high risk primary breast cancer. J Clin Oncol 1993; 11: 1132–1143.[Abstract]

8. Hortobagyi GN, Buzdar AU, Theriault RL et al. Randomised trial of high-dose chemotherapy and blood cell autografts for high-risk primary breast carcinoma. J Natl Cancer Institute 2000; 92: 3225.

9. Rodenhuis S, Richel DJ, Van der Wall E et al. Randomised trial of high-dose chemotherapy and haemopoietic progenitor-cell support in operable breast cancer with extensive axillary lymph-node involvement. Lancet 1998; 352: 515–521.[CrossRef][ISI][Medline]

10. Bergh J, Wiklund T, Erikstein B et al. Tailored fluorouracil, epirubicin, and cyclophosphamide compared with marrow-supported high-dose chemotherapy as adjuvant treatment for high-risk breast cancer: a randomised trial. Lancet 2000; 356: 1384–1391.[CrossRef][ISI][Medline]

11. Coombes RC, Bliss JM, Wils J et al. Adjuvant Cyclophosphamide, methotrexate, and Fluorourcil versus Fluorouracil, epirubicin and cyclophosphamide chemotherapy in pre-menopausal women with axillary node-positive operable breast cancer: Results of a Randomised Trial. J Clin Oncol 1996; 24: 35–45.

12. Sayer HG, Will U, Schilling K et al. Hepatic veno-occlusive diseased (VOD) with complete occlusion of liver venules after tandem autologous stem cell transplantation—successful treatment with high-dose methylprednisolone and defibrotide. J Cancer Res Clin Oncol 2003; 128: 148–152.[CrossRef][ISI]

13. Cox DR. Regression models and life tables. J R Stat Soc 1972; 34: 187–220.[ISI]

14. Rodenhuis S, Bontenbal M, Beex L et al. High-dose chemotherapy with haematopoetic stem-cell rescue for high-risk breast cancer. N Engl J Med 2003; 349: 7–16.[Abstract/Free Full Text]

15. Couch RD, Loh KK, Sugino J. Sudden cardiac death following adriamycin therapy. Cancer 1981; 48: 38–39.[ISI][Medline]

16. Peters WP, Rosner G, Vredenburgh J et al. Updated results of a prospective, randomised comparison of two doses of combination alkyating agents (AA) as consolidation after CAF in high-risk primary breast cancer involving ten or more axillary lymph nodes (LN): CALGB 9082/SWOG 9114/NCIC Ma-13. ASCO 2001; 81: (Abstr).

17. Leonard RCF, Lind M, Twelves C et al. Conventional adjuvant chemotherapy versus single cycle, autograft-supported, high dose, late-intensification chemotherapy in high-risk breast cancer patients: a randomized trial. J Natl Cancer Inst 2004, 96; 1076–1083.[Abstract/Free Full Text]

18. Roche HH, Pouillart P, Meyer N et al. Adjuvant high dose chemotherapy (HDC) improves early outcome for high risk (N > 7) breast cancer patients: the Pegase 01 trial. ASCO 2001; 102: (Abstr).

19. Gianni A, Bonadonna G. Five-year results of the randomised clinical trial comparing standard versus high-dose myeloablative chemotherapy in the adjuvant treatment of breast cancer with >3 positive nodes (LN+). ASCO 2001; 80: (Abstr).

20. High-dose chemotherapy with haematopoietic stem cell transplantation for metastatic breast cancer patients: final results of the French multicentric randomised CMA/PEGASE 04 protocol. Eur J Cancer. 2005; 41: 71–80.[CrossRef][ISI][Medline]

21. Yau JC, Gertler SZ, Hanson J et al. A phase III study of high-dose intensification without hematopoietic progenitor cells support for patients with high-risk primary breast carcinoma. J Clin Oncol 2000; 23: 292–296.

22. Farquhar C, Basser R, Marjoribanks J et al. High dose chemotherapy and autologous bone marrow or stem cell transplantation versus conventional chemotherapy for women with early poor prognosis breast cancer (Cochrane Review). The Cochrane Library, Issue 2003; 1.

23. Tallman MS, Gray R, Robert NJ et al. Conventional adjuvant chemotherapy with or without high-dose chemotherapy and autologous stem-cell transplantation in high-risk breast cancer. N Engl J Med 2003; 349: 1.[Free Full Text]