A randomised phase II study of conventional versus accelerated infusional chemotherapy with granulocyte colony-stimulating factor support in advanced breast cancer

R. H. De Boer1, T. G. Eisen1, P. A. Ellis2, S. R. D. Johnston1, G. Walsh1, S. Ashley1 and I. E. Smith1,+

1The Royal Marsden NHS Trust, London; 2Guys and St Thomas’s NHS Trust, London, UK

Received 26 March 2001; revised 10 December 2001; accepted 9 January 2002.


    Abstract
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Background

Granulocyte colony-stimulating factor (G-CSF) allows cycles of conventional bolus chemotherapy to be accelerated with reduction in treatment time and a boost in dose intensity. Theoretically, this approach could be hazardous with infusional 5-fluorouracil (5-FU) chemotherapy, since G-CSF-stimulated neutrophil proliferation would be occurring in the face of continuous S-phase active 5-FU. We performed this phase II randomised study to compare the safety, tolerability and efficacy of conventional 3-weekly epirubicin, cyclophosphamide and continuous infusional 5-FU (infusional ECF) to an accelerated 2-weekly schedule with G-CSF support, in patients with advanced breast cancer.

Patients and methods

Twenty-seven patients were randomised, with 14 in the accelerated arm. Patients received bolus epirubicin 60 mg/m2 and cyclophosphamide 600 mg/m2 every 3 weeks (conventional arm) or every 2 weeks (accelerated arm) and 5-FU 200 mg/m2/day continuous infusion throughout. G-CSF 300 µg/day s.c. on days 10–12 was given each accelerated cycle.

Results

There were no treatment delays secondary to inadequate neutrophil or platelet recovery in either arm, with higher median day 1 neutrophil counts for each cycle in the accelerated arm compared with the conventional arm. Eighty-six per cent of the planned conventional chemotherapy cycles and 82% of the planned accelerated cycles were given. There were no major differences in toxicity between the arms, with the most common grade 3 toxicities being alopecia and stomatitis. Eight patients developed neutropenic sepsis (five in the accelerated arm and three in the conventional arm). Ten patients (77%) responded in the conventional arm and nine (64%) in the accelerated arm.

Conclusions

Accelerated infusional ECF with limited G-CSF support is a feasible and well-tolerated regimen with rapid haematological recovery. A 50% increase in relative dose intensity of epirubicin and cyclophosphamide is achieved, while overall treatment time is reduced by 33%.

Key words: accelerated, breast cancer, cyclophosphamide, epirubicin, 5-fluorouracil, granulocyte colony-stimulating factor


    Introduction
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Several studies in patients with advanced breast cancer have investigated accelerating chemotherapy cycles from the standard 3 or 4 weeks to 2 weeks with haematopoietic growth factor support. The majority of these studies have found that the approach is both feasible and safe with reported response rates of up to 87% and increases in dose intensity of up to 50% [17]. The type of growth factor used [granulocyte colony-stimulating factor (G-CSF) or granulocyte–macrophage colony-stimulating factor (GM-CSF)], the dose administered and the schedule has varied widely. Accelerated chemotherapy schedules have also been applied in the neo-adjuvant and the high-risk adjuvant settings [810]. At the Royal Marsden Hospital, we developed a chemotherapy schedule of continuous infusional 5-fluorouracil (5-FU) with 3-weekly bolus epirubicin and cyclophosphamide (infusional ECF regimen), initially for patients with advanced disease [11], then inflammatory breast cancer [12] and then in patients with large operable disease [13]. The high activity of this regimen encouraged us to test whether this regimen could be accelerated to a 2-weekly schedule to decrease the time required for patients to complete their treatment programme and possibly also to improve efficacy. The theoretical concern of accelerating this schedule with growth factor support was the potential interaction between G-CSF and infusional 5-FU. G-CSF causes stimulation and increased proliferation of haematopoietic progenitor cells in the bone marrow with a resultant increase in the number of circulating neutrophils [14]. Exposure of these rapidly proliferating cells to an S-phase specific cytotoxic agent could potentially lead to significant cellular death, with subsequent failure of neutrophil recovery and prolongation of neutropenia. De Witt et al. [15] reported that this was more than just a theoretical concern with their findings that the administration of bolus chemotherapy after the administration of G-CSF worsened the toxic effects of chemotherapy on the bone marrow.

With this in mind, we conducted this phase II randomised study to determine the safety, tolerability and effectiveness of an accelerated infusional ECF schedule in patients with locally advanced or metastatic breast cancer.


    Patients and methods
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Patient selection
Patients with a histological diagnosis of breast cancer with either measurable metastatic disease or locally advanced (inoperable) disease (including patients with inflammatory breast cancer) were eligible. Other eligibility criteria included good performance status [World Health Organization (WHO) grade 0–1], adequate renal function (EDTA clearance >60 ml/min), cardiac function (left ventricular ejection fraction >50%), hepatic function [bilirubin (<30 µmol) and hepatic enzymes <2 x upper limit of normal] and bone marrow function (white blood cells >3.0 x 109/l; platelets >100 x 109/l). Although prior chemotherapy or endocrine therapy for either adjuvant or metastatic disease was allowed, prior treatment with either an anthracycline or with infusional 5-FU was not allowed. Prior exposure to bolus 5-FU was acceptable. The study protocol was approved by the hospital ethics committee and all patients gave written informed consent.

Methods
Initial work-up included full history and physical examination, with two-dimensional tumour measurement in patients with locally advanced/inflammatory disease, full blood count, plasma urea and electrolytes, liver function tests, 12-lead electrocardiogram, chest X-ray, isotopic bone scans and liver imaging [ultrasound or computed tomography (CT) scan]. Further CT/magnetic resonance imaging scanning was only performed if there was a clinical or biochemical suspicion of metastatic disease at other sites. All patients had a multiple gated acquisition scan to assess left ventricular function.

Treatment
The conventional epirubicin, cyclophosphamide and continuous infusional 5-FU (infusional ECF) chemotherapy schedule consisted of a continuous infusion of 5-FU administered at a dose of 200 mg/m2 every 24 h via an ambulatory infumed pump and a Hickman line inserted into a subclavian vein for a planned 18 weeks. This was combined with bolus intravenous injections of epirubicin (60 mg/m2) and cyclophosphamide (600 mg/m2), given once every 3 weeks for a planned six courses.

The accelerated infusional ECF chemotherapy schedule consisted of a continuous infusion of 5-FU administered at a dose of 200 mg/m2 every 24 h as described above, for a planned 12 weeks. This was combined with bolus epirubicin (60 mg/m2) and cyclophosphamide (600 mg/m2) as described above, given once every 2 weeks for a planned six courses.

Antiemetic cover was provided with either ondansetron 8 mg i.v. or granisetron 3 mg i.v. combined with dexamethasone 8 mg i.v., followed by 3 days of oral dexamethasone 4 mg three times daily and domperidone 20 mg four times daily. Scalp cooling was offered with each course of epirubicin. All patients were treated with low dose warfarin (1 mg/day) to prevent Hickman line-related thrombosis. G-CSF was given subcutaneously at a dose of 300 µg/day on days 10–12 to all patients in the accelerated arm. Longer schedules of G-CSF were allowed if deemed necessary for patient safety by the treating physician.

Assessment of response and toxicity
Patients were assessed clinically every cycle with bidimensional tumour measurements. Overall response rates to chemotherapy were determined using standard WHO criteria [16]. Toxicity for each cycle was recorded prior to the commencement of the following cycle and was also assessed using the standard WHO criteria [16]. Nadir blood counts were not routinely taken in patients receiving conventional ECF.

Dose intensity calculation
Dose intensity was calculated for individual patients for all received treatment, expressed for each drug in mg/m2/week.

Dose modifications
Dose modifications were made for toxicity as follows.

Myelosuppression
If the total neutrophil count was <1.0 x 109/l and/or the platelet count was <100 x 109/l, the 5-FU was continued but the epirubicin and the cyclophosphamide were delayed for a week. If longer delays were required, then dose reductions of all drugs occurred.

Palmar-plantar syndrome
Pyridoxine 50 mg orally three times per day was started if symptoms were mild, and 5-FU infusion was discontinued if symptoms were more severe. Upon healing, 5-FU infusion was recommenced with a 25% dose reduction.

Diarrhoea
For severe/persistent diarrhoea, 5-FU was discontinued for 1 week and restarted with a 25% dose reduction.

End points and statistical considerations
It was planned to use a two stage Fleming’s phase II design [17]. In the first stage, toxicity levels were assessed after 10 patients had been treated in the accelerated arm. Myelosuppression of grade 2 or higher in 90% of patients was deemed unacceptable and the trial was assessed after 10 patients so that it could be stopped early if such high levels of toxicity were seen. If all 10 patients showed grade >=2 myelosuppression then the study would be discontinued. In the second stage, randomisation was planned to continue until a further 5–10 patients had been treated on each arm. A one-to-one randomisation was done using a permuted block design (block size 10). Then, if 10 or more patients on the accelerated arm showed tolerable levels of myelosuppression (grade <=1), consideration would be given to expanding to a full phase III trial ({alpha} = 4%; power = 95%).


    Results
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Twenty-seven patients with locally advanced or metastatic breast cancer were randomised into this study at the Royal Marsden Hospital, London. Patient demographics are shown in Table 1.


View this table:
[in this window]
[in a new window]
 
Table 1. Patient demographics
 
Treatment delivered
Eight-six per cent and 82% of the total planned treatment cycles were administered to patients in the conventional and accelerated arms, respectively, with the median cycles per patient being the same (five cycle, range 1–6). Nine patients in the conventional arm (69%) and eight in the accelerated arm (57%) received all six planned cycles of treatment. There was no difference in the number of patients requiring treatment delays (three versus one) or in ceasing treatment early due to toxicity (three versus four). Five patients in each arm required dose reductions during treatment, with another two patients in the accelerated arm having their initial epirubicin dose reduced by 25% due to tumour-related liver function abnormalities. The majority of the dose reductions were of 5-FU in response to mucous membrane or skin toxicities. This resulted in a slight decrease in delivered dose intensity of 5-FU compared with the delivered dose intensity of epirubicin and cyclophosphamide. With respect to the dose intensity analysis, the doses and time interval in the conventional regimen were assigned a relative dose intensity (RDI) of 1.0. The delivered dose intensity of the epirubicin and cyclophosphamide in the accelerated arm was increased (1.47 and 1.47 RDI, respectively) compared with the conventional arm (0.99 and 0.99 RDI, respectively). The majority of patients in the accelerated arm received G-CSF injections as planned (10 of 14, 71%). Four patients (29%) were given longer schedules of G-CSF; one for a single cycle and three for multiple cycles. This was in response to neutropenic infections.

Toxicity
Haematological
Haematological toxicity is shown in Table 2. This demonstrates the positive impact of G-CSF on haematological recovery. Despite day 1 in the accelerated arm being 7 days earlier than day 1 in the conventional arm, the median neutrophil counts were consistently higher in patients receiving accelerated treatment plus G-CSF. Neutrophil counts of up to 21.3 x 109/l were seen, and the impact of G-CSF did not appear to diminish as the patients progressed through the treatment cycles, with median levels remaining high throughout. There was no major difference between the two treatment arms in the number of patients developing neutropenic sepsis (five patients in the accelerated arm versus three in the conventional arm) although the overall rate of eight of 27 (30%) appeared high. The five patients in the accelerated arm developed neutropenic sepsis between days 8 and 11 post-chemotherapy in cycles 1, 2, 2, 4 and 5, respectively. All were admitted to hospital for intravenous antibiotics and stayed for a median of 3.5 days (range 2–5 days). The median length of neutropenia was 2.5 days (range 1–4 days). The three patients in the conventional arm developed neutropenic sepsis on day 10 cycle 1, day 11 cycle 1 and day 16 cycle 5, respectively. All were admitted to hospital for intravenous antibiotics and stayed for 2–3 days with neutrophil recovery occurring within 1–2 days.


View this table:
[in this window]
[in a new window]
 
Table 2. Median neutrophil, platelet and haemoglobin counts on scheduled day 1 of each cycle
 
There appeared to be no major impact of the accelerated schedule on platelet counts (Table 2). In both arms there was a trend towards lower haemoglobin levels as the number of administered treatment cycles increased, with the effect seen in the accelerated arm being somewhat more pronounced (Table 2). Nevertheless, anaemia of greater than grade 2 in severity was not seen.

Non-haematological
The major non-haematological toxicities experienced by patients in each treatment arm are summarised in Table 3. In general, both schedules were well tolerated, with no WHO grade 4 toxicities experienced.


View this table:
[in this window]
[in a new window]
 
Table 3. Non-haematological toxicity (worst grade)a
 
Treatment outcomes
Overall clinical responses were similar between the two groups [10 (77%) in the conventional arm and nine (64%) in the accelerated arm]. Five of 14 patients (36%) in the accelerated arm achieved a complete clinical response (CR) compared with two of 13 (15%) in the conventional arm. One patient in the accelerated arm developed progressive disease after two cycles of treatment. This patient had multiple liver metastases at presentation, and required an initial epirubicin dose reduction due to abnormal liver function.

The Eastern Cooperative Oncology Group performance status of each patient was recorded on the first day of each cycle. The majority of the patients in both arms had either stable or improved performance status during their treatment. Only three patients suffered a deterioration, and of these two were temporary, being associated with an acute toxicity.

The median time to progression was similar in the two groups [12 months (range 1–38) in the conventional arm and 10.6 months (range 1–25) in the accelerated arm], as was the median overall survival [18.3 months (range 1–41) and 19.3 months (range 3–38), respectively]. Still alive at the time of analysis were two patients in the conventional arm (at 28 and 40 months follow-up) and three in the accelerated arm (at 20, 25 and 32 months follow-up).


    Discussion
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
This randomised phase II study shows that accelerating conventional 3-weekly infusional ECF into a 2-week schedule with limited G-CSF support is feasible and tolerable in patients with advanced breast cancer, without any increase in haematological toxicity that might have been predicted on theoretical grounds. A number of previous studies have demonstrated that conventional chemotherapy regimens can be accelerated to a 2-weekly schedule with growth factor support [17]. To our knowledge, this is the first study to add a haematopoietic growth factor to a schedule containing continuous infusional 5-FU. Prior to this study concerns had been raised that stimulation of haematopoietic precursor cells in the face of continuous exposure to a cell cycle-specific cytotoxic agent might actually have a detrimental impact upon neutrophil recovery and increase or prolong neutropenic episodes. The study in fact showed that neutrophil recovery was rapid and complete in the accelerated treatment arm with higher median counts on day 1 of each treatment cycle than seen in patients in the conventional arm. None of the 14 patients receiving accelerated treatment required a dose delay due to inadequate neutrophil counts on day 1 of their next scheduled treatment cycle. Thus, growth factor support in this setting is both feasible and effective.

The 3-day schedule of G-CSF administration (from days 10 to 12) was empirically chosen, based upon local experience and cost considerations. Although standard schedules of G-CSF administration often begin on the first day post-chemotherapy and continue for up to 10 days, the data supporting such prolonged schedules in patients with solid tumours receiving standard doses of chemotherapy is weak. Papaldo et al. [18] reported a randomised study of accelerated chemotherapy in advanced breast cancer with patients receiving G-CSF either from days 3 to 7, or from days 3 to 12. Although there were only 22 patients in the study, there was no difference in haematological toxicity between the two arms. Four of 14 patients in our accelerated arm received a longer schedule of G-CSF (one for a single cycle and three for multiple cycles). Nevertheless, the majority of patients had excellent neutrophil recovery with 3 days of administration. This shorter schedule of G-CSF has advantages in terms of both cost and simplification of treatment. Of note, a total of eight patients developed febrile neutropenia during study treatment. These patients required hospitalisation and treatment with intravenous antibiotics. The patients were equally distributed between the two treatment groups (five in the accelerated arm versus three in the conventional arm), with two of the patients in the accelerated arm having open wounds at the time of randomisation (one had a large fungating untreated breast lesion, and one had an open excisional breast biopsy site). It is uncertain whether a longer schedule of G-CSF administration would have decreased the incidence of febrile neutropenia. Non-haematological treatment toxicities were largely as predicted, with both arms well tolerated by the majority of patients.

There were five clinical complete remissions seen in the accelerated arm (36%) versus two in the standard arm (15%). These findings are consistent with previous studies of accelerated treatment which found high rates of complete clinical remissions, ranging up to 30% of patients treated [1, 6, 19, 20]. However, the small total number of patients studied makes it impossible to determine whether the 48% increase in the relative dose intensity of epirubicin and cyclophosphamide results in an increase in response rate or in improved survival outcomes.

Although important, improved disease outcomes are only one aim of accelerated therapy. Another consideration is the shortened treatment time they offer, thus reducing the impact of treatment on the lives of the patients. In this study, palliative therapy could be administered over a 12-week period rather than over 18 weeks, a reduction of 33% in treatment time. Whether this in itself makes accelerated treatment worthwhile is a value judgement, particularly in view of the extra cost involved with the use of G-CSF and the slight increase in grade 3 toxicities and febrile neutropenia rate. In addition, it remains unclear whether shorter treatment schedules per se may have a detrimental impact upon treatment outcomes in the metastatic setting.

This study has demonstrated that accelerated infusional ECF chemotherapy is feasible and tolerable in patients with locally advanced or metastatic breast cancer. Acceleration of infusional ECF increases RDI by up to 50% whilst reducing treatment time by 33%. The development of oral 5-FU agents such as eniluracil [21] and capecitabine [22] in the management of advanced breast cancer raises the possibility of replacing the infusional component of the treatment schedule with daily oral dosing, therefore eliminating the need for invasive catheters and pumps. A larger phase III study is required to determine fully the potential benefits of this regimen.


    Footnotes
 
+ Correspondence to: Professor I. E. Smith, Department of Medicine, The Royal Marsden NHS Trust, Fulham Road, London SW3 6JJ, UK. Tel: +44-207-808-2751; Fax: +44-207-352-5441; E-mail: ian.smith@rmh.nthames.nhs.uk Back


    References
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
1. Ardizzoni A, Venturini M, Setoli MR et al. Granulocyte-macrophage colony-stimulating factor (GM-CSF) allows acceleration and dose-intensity increase of CEF chemotherapy: A randomised study in patients with advanced breast cancer. Br J Cancer 1994; 69: 385–391.[ISI][Medline]

2. Bissett D, Jodrell D, Harnett AN et al. Phase I study of accelerated FEC with granulocyte colony-stimulating factor (Lenograstim) support. Br J Cancer 1995; 71: 1279–1282.[ISI][Medline]

3. Piccart MJ, Bruning P, Wilders J et al. An EORTC pilot study of filgrastim (recombinant human granulocyte colony stimulating factor) as support to a high dose-intensive epiadriamycin-cyclophosphamide regimen in chemotherapy-naïve patients with locally advanced or metastatic breast cancer. Ann Oncol 1995; 6: 673–677.[Abstract]

4. Scinto AF, Ferraresi V, Campioni N et al. Accelerated chemotherapy with high-dose epirubicin and cyclophosphamide plus r-met-HUG-CSF in locally advanced and metastatic breast cancer. Ann Oncol 1995; 6: 665–671.[Abstract]

5. Lalisang RI, Wils JA, Nortier HW et al. Comparative study of dose escalation versus interval reduction to obtain dose-intensification of epirubicin and cyclophosphamide with granulocyte colony-stimulating factor in advanced breast cancer. J Clin Oncol 1997; 15: 1367–1376.[Abstract]

6. Nistico C, Garufi C, Barni S et al. Phase II study of epirubicin and vinorelbine with granulocyte colony-stimulating factor: a high-activity, dose-dense weekly regimen for advanced breast cancer. Ann Oncol 1999; 10: 937–942.[Abstract]

7. Cottu PH, Zelek L, Extra JM et al. High-dose epirubicin and cyclophosphamide every two weeks as first-line chemotherapy for relapsing metastatic breast cancer patients. Ann Oncol 1999; 10: 795–801.[Abstract]

8. Del Mastro L, Garrone O, Sertoli MR et al. A pilot study of accelerated cyclophosphamide, epirubicin and 5-fluorouracil plus granulocyte colony stimulating factor as adjuvant therapy in early breast cancer. Eur J Cancer 1994; 30A: 606–610.

9. Lorusso V, Catino A, Schittulli F et al. Neoadjuvant chemotherapy with accelerated CNF plus G-CSF in patients with breast cancer tumours larger than three centimetres: A pilot study. Int J Oncol 1998; 12: 1177–1181.[ISI][Medline]

10. Fountazilias G, Nicolaides C, Aravantinos G et al. Dose-dense adjuvant chemotherapy with epirubicin monotherapy in patients with operable breast cancer and >=10 positive axillary lymph nodes. A feasibility study. Oncology 1998; 55: 508–512.[ISI][Medline]

11. Jones AL, Smith IE, O’Brien MER et al. Phase II study of continuous infusional 5-FU with epirubicin and cisplatin (infusional ECF) in patients with metastatic and locally advanced breast cancer: An active new regimen. J Clin Oncol 1994; 12: 1259–1265.[Abstract]

12. de Boer RH, Saini A, Johnston SRD et al. Continuous infusional combination chemotherapy in inflammatory breast cancer: A phase II study. Breast J 2000; 9: 149–155.

13. Smith IE, Walsh G, Jones A et al. High complete remission rates with primary neoadjuvant chemotherapy for large early breast cancer. J Clin Oncol 1995; 13: 424–429.[Abstract]

14. Molineux G, Dexter TM. Biology of G-CSF. In Morstyn G, Dexter TM (eds): Filgrastim in Clinical Practice. New York, NY: Marcel Dekker 1994; 1–23.

15. De Wit R, Verweij J, Bontenbal M et al. Adverse effect on bone marrow protection of pre-chemotherapy granulocyte colony-stimulating factor support. J Natl Cancer Inst 1996; 88: 1393–1398.[Abstract/Free Full Text]

16. Miller AB, Hoogstraten B, Straquet M, Winkler A. Reporting results of cancer treatment. Cancer 1981; 147: 207–214.

17. Fleming TR. One sample multiple testing procedure for phase II clinical trials. Biometrics 1982; 38: 143–151.[ISI][Medline]

18. Papaldo P, Marolla P, Pellegrini F, Calabresi F. Accelerated chemotherapy (CT) with two schedules of granulocyte colony-stimulating factor (G-CSF) in advanced breast cancer. Ann Oncol 1992; 3 (Suppl 5): 7.

19. Ries F, Duhem C, Kleiber K, Dicato M. Phase I/II clinical trial of epirubicin and paclitaxel followed by granulocyte colony-stimulating factor in a 2-week schedule in patients with advanced or metastatic breast cancer. Semin Oncol 1997; 24: (Suppl 17): 48–51.[ISI]

20. Stoger H, Sammonigg H, Krainer M et al. Dose intensification of epidoxorubicin and cyclophosphamide in metastatic breast cancer: a randomised study with two schedules of granulocyte-macrophage colony stimulating factor. Eur J Cancer 1998; 34: 482–488.[ISI][Medline]

21. Smith IE, Johnson SRD, O’Brien MER et al. Low-dose fluorouracil with eniluracil as first-line chemotherapy against advanced breast cancer: a phase II study. J Clin Oncol 2000; 18: 2378–2384.[Abstract/Free Full Text]

22. Blum JC, Jones SE, Buzdar AU et al. Multicentre phase II study of capecitabine in paclitaxel-refractory metastatic breast cancer. J Clin Oncol 1999; 17: 485–493.[Abstract/Free Full Text]





This Article
Abstract
Full Text (PDF)
E-letters: Submit a response
Alert me when this article is cited
Alert me when E-letters are posted
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (1)
Disclaimer
Request Permissions
Google Scholar
Articles by De Boer, R. H.
Articles by Smith, I. E.
PubMed
PubMed Citation
Articles by De Boer, R. H.
Articles by Smith, I. E.