- © 2002 by American Society of Clinical Oncology
Phase II Study of Oxaliplatin and Fluorouracil in Taxane- and Anthracycline-Pretreated Breast Cancer Patients
- L. Zelek,
- P. Cottu,
- M. Tubiana-Hulin,
- J.-M. Vannetzel,
- Ph. Chollet,
- J.-L. Misset,
- N. Chouaki,
- M. Marty,
- E. Gamelin,
- S. Culine,
- V. Dieras,
- S. Mackenzie and
- M. Spielmann
- From the Institut Gustave Roussy; Hôpital Paul Brousse; Cvitkovic et Associés Consultants, Villejuif; Hôpital Saint-Louis; Institut Curie, Paris; Centre René Huguenin, St Cloud; Clinique Hartmann, Neuilly sur Seine; Centre Jean Perrin, Clermont Ferrand; Centre Paul Papin, Angers; and Centre Régional de Lutte Contre le Cancer Val d’Aurelle, Montpellier, France.
- Address reprint requests to M. Spielmann, MD, Institut Gustave Roussy, 39 rue Camille Desmoulins, 94805 Villejuif, France; email: spielman{at}igr.fr
Abstract
PURPOSE: Phase II study evaluating efficacy and safety of combined oxaliplatin/fluorouracil (5-FU) in taxane-pretreated advanced and metastatic breast cancer (ABC) patients.
PATIENTS AND METHODS: Sixty-four taxane- and anthracycline-pretreated (within 6 months of study entry) women were treated with oxaliplatin 130 mg/m2 (2-hour intravenous [IV] infusion), day 1, and 5-FU 1,000 mg/m2/d (continuous IV infusion) days 1 to 4, every 3 weeks.
RESULTS: Median patient age was 51 years (range, 34 to 71 years), with a median of two involved organs (range, one to six organs), and metastases in the liver (70%), bone (47%), and lung (34%). Patients had a median of two prior chemotherapy regimens (range, one to six regimens), and 78% had previous hormonal therapy, with clinical taxane and anthracycline resistance in 53% and 34%, respectively. A total of 367 cycles were administered, with a median of six cycles/patient (range, one to 15 cycles). Sixty patients were assessable for response (World Health Organization criteria): 17 partial response, 26 stable disease, and 17 disease progression, giving an overall response rate of 27% (95% confidence interval, 16.3% to 39.1%), and 26% and 36% in taxane- and anthracycline-resistant populations, respectively, all responders having metastatic liver disease. Median time to progression was 4.8 months, and median overall survival was 11.9 months. Four treatment-related serious adverse events occurred, seven patients withdrew because of treatment-related toxicity. Hematotoxicity was prevalent but rarely severe, with grade 3-4 neutropenia, leukopenia, and thrombocytopenia in 34%, 19%, and 16% of patients, respectively, and a single episode of febrile neutropenia. One third of patients developed grade 2-3 peripheral neuropathy (oxaliplatin-specific scale), with grade 3 in only 8%.
CONCLUSION: This oxaliplatin/5-FU combination is effective with an excellent safety profile in anthracycline/taxane-pretreated ABC patients, showing encouraging activity in patients with anthracycline/taxane-resistance or visceral disease.
BREAST CANCER IS the most common cancer affecting women in industrialized Western countries. Overall, it represents 9.4% of all reported cases of cancer and is responsible for one third of all female cancer deaths.1 Hormonal and/or cytotoxic treatment has been increasingly used in breast cancer patients,2 and studies have shown combination chemotherapy to be superior to single-agent treatment.2,3 However, although objective tumoral responses are initially achieved in approximately two thirds of patients with advanced or metastatic disease, complete response in all sites is obtained in no more than 10% of cases, and the median duration of survival rarely exceeds 2 years.2,4
Many cytotoxic agents have demonstrated activity in advanced breast cancer (ABC), the most active and widely used being anthracyclines, taxanes, fluorouracil (5-FU), cyclophosphamide, mitomycin, and vinorelbine, which are generally given in combination treatments. Anthracyclines are considered the reference chemotherapy for ABC. Doxorubicin is the most commonly used agent, giving an objective response rate close to 50% when given as first-line therapy.5 Higher response rates (55% to 80%) and a longer median duration of response are seen when anthracyclines are combined with other agents.6 The use of anthracyclines in the adjuvant setting, either as single agents or in combination, has become more prevalent since its initial use7 because of the therapeutic advantage in this clinical setting.7,8 However, their use in metastatic settings has been limited by their cumulative cardiotoxicity,9 notably in the treatment of anthracycline pretreated patients at an early stage of disease. Furthermore, anthracycline-pretreated patients who receive adjuvant anthracyclines as second-line therapy have lower response rates and a shorter duration of response than anthracycline-naive patients receiving adjuvant anthracyclines.
Among the taxanes, paclitaxel has been used in the treatment of breast cancer, showing efficacy, a good toxicity profile, and a lack of cross-resistance with anthracyclines. Used in first-line therapy in metastatic disease, several studies have reported overall response rates of 30% to 60%, and 20% to 40% when used as second-line or salvage single-agent therapy in this population.10 Recent trials have shown that docetaxel is superior to doxorubicin.11 The recommended dose of docetaxel has elicited the highest reported response rate for any single-agent chemotherapy in first-line treatment. This has been reported for patients with metastatic disease,12 for anthracycline-pretreated refractory patients, and for patients with visceral disease.13 The therapeutic value of docetaxel over other drugs whose use is already established in ABC chemotherapy is principally in its high activity in visceral metastatic sites, notably in liver metastases, and its sustained activity in anthracycline-resistant patients.
The combination of taxanes with anthracyclines has rapidly gained acceptance in ABC treatment, with response rates and time-related parameters consistently higher than those previously reported.14 In addition, the use of taxane/anthracycline combinations both in first-line metastatic and in large adjuvant programs is rapidly increasing, creating a need to develop new options after failure of taxane/anthracycline adjuvant combinations or in patients at risk of developing cumulative dose toxicities (eg, cardiotoxicity and fluid retention). In addition, the high prevalence of pleiotropic clinical resistance with multiple drug resistance proteins and other overexpressed cell membrane pumps, changes in apoptotic regulation, alterations to DNA repair, and mutations in topoisomerase II and tubulin, contribute to poor responsiveness to both anthracyclines and taxanes.
Both the thymidylate synthase inhibitor 5-FU and, more recently, its oral prodrug capecitabine (registered in Europe as of March 2001) have proven efficacy when administered as palliative therapy in breast cancer patients who have exhausted anthracycline and taxane therapeutic options.15,16 However, activity, although consistent, is seen in less than one third of these patients and is short-lasting.
Oxaliplatin, a diaminocyclohexane platinum with a large preclinical spectrum of anticancer activity, has been recently approved in Europe and elsewhere for use in colorectal cancer patients in combination with 5-FU. It forms intrastrand DNA adducts,17,18 which differ from those of cisplatin or carboplatin by their repair capability. Oxaliplatin-induced DNA damage has a higher cytotoxic potency and is active in mismatch repair deficient cells, accounting for its activity in cisplatin-resistant cell lines and cancer patients.19 Oxaliplatin has a different spectrum of preclinical activity than other platinum compounds, which is linked to genetic and epigenetic mismatch repair deficiencies. Its safety profile is very favorable, limited only by mild hematotoxicity and cumulative neurosensory toxicity, the latter usually being reversible on discontinuation.20,21
There is a clear need for feasible therapeutic options for ABC patients who have failed treatment with anthracyclines and/or taxanes. A 5-FU/oxaliplatin combination regimen was investigated for several reasons. First, in a pilot phase II clinical study conducted in 14 anthracycline-resistant female ABC patients treated with 130 mg/m2 oxaliplatin every 3 weeks, three partial responses (23%), one stable disease, and nine progressive diseases were observed, with an acceptable toxicity profile.22 In addition, in vitro studies have shown that oxaliplatin has a high level of cytotoxic activity in several breast cancer cell lines of human origin, and antitumoral efficacy in vivo in the hormone-refractory breast cancer GR-mouse mammary model, which mimics hormone-refractory human breast cancer.23 Although cisplatin and carboplatin are active agents in previously untreated ABC,24,25 they have no relevant activity in pretreated patients, which is most likely because of increased functional mismatch repair deficit in these pretreated patients occurring after exposure to alkylators or intercalators.26 However, unlike cisplatin or carboplatin, oxaliplatin is active in mismatch-repair deficient cells, a condition that is prevalent in ABC and dramatically increases after exposure to anthracycline-containing chemotherapy.26 Secondly, 5-FU, given as a continuous infusion for metastatic or locally advanced breast cancer, shows efficacy in adjuvant, neoadjuvant, and palliative settings.27 Thirdly, oxaliplatin and 5-FU have shown synergy in preclinical studies,18 which has been confirmed in clinical studies in colorectal cancer patients.28 Furthermore, using the GR breast cancer murine animal model, high potentiation of oxaliplatin efficacy by 5-FU has been observed in vitro and in vivo.23 Finally, this combination has a favorable toxicity profile, which is supported by extensive clinical evaluation in patients with advanced colorectal cancer.28-30 The added value of infusional 5-FU over bolus modality has been validated in colorectal and head and neck cancers; whereas both its recent revival in pretreated ABC31 and the activity of capecitabine, an oral continuous intravenous (IV) infusion,32 support its choice for combination therapy in this disease. Thus, this study was designed to investigate the activity and toxicity profile of an oxaliplatin/5-FU combination in taxane- and anthracycline-pretreated ABC patients.
PATIENTS AND METHODS
Patient Population
All eligible patients had to meet the following criteria: (1) they had to have advanced or metastatic, histologically or cytologically proven breast cancer; (2) at least one uni- or bidimensionally measurable lesion that had been measured less than 4 weeks before inclusion; (3) received conventional taxane treatment no more than 6 months before study entry; (4) documented disease progression under their last treatment; (5) not received chemotherapy in the 4 weeks before study entry; (6) not received continuous 5-FU and/or platinum therapy in the 12 months before study entry or 5-FU bolus therapy in the 6 months before study entry; (7) a World Health Organization performance status ≤ 2; (8) they needed to be between the ages of 18 and 75 years; (9) they had to have a life expectancy ≥ 3 months; and (10) adequate hepatic, renal, and bone marrow function were required.
Patients were excluded from the study for any of the following reasons: (1) if they were experiencing symptomatic peripheral neuropathy National Cancer Institute common toxicity criteria grade more than 1; (2) were pregnant or breast-feeding; (3) had a history of prior malignancies (with the exception of excised cervical carcinoma-in-situ or nonmelanoma cell skin carcinoma); (4) were receiving or had received, in the 30 days before study screening, any treatment with experimental drugs; (5) had known brain or leptomeningeal involvement; (6) had a serious medical condition; and (7) were receiving concomitant treatment with any other anticancer therapy, with corticosteroids other than as antiemetic premedication, or with biphosphonates except in cases of chronic use. The protocol was approved by the institutional ethics committee, and signed informed consent was obtained from all patients. Taxane and anthracycline clinical resistance were defined as disease progression occurring during the relevant previous treatment including the 4 weeks after the last treatment administration (with a third party, medically-verified assessment).
Study Design
This was a multicenter phase II trial in which patients received 130 mg/m2 oxaliplatin in 5% glucose, administered as a 2-hour IV infusion, and 1,000 mg/m2/d 5-FU as a continuous IV infusion over 4 days, every 3 weeks. Antiemetic treatment included 5-hydroxytryptamine-3 receptor antagonists, with any other symptomatic management being up to the investigator. Patients received medication for pain management as required, including antiemetics, benzodiazepine, and so on.
Pretreatment evaluation included a complete medical history, physical examination, chest x-ray, chest and abdominal computed tomography scan, and computed tomography scan of all measurable/assessable sites. Complete differential blood cell counts were performed and blood chemistry parameters were measured at baseline and before each cycle. Complete blood cell counts were repeated weekly. The tumor markers carcino-embryonic antigen (CEA) and Ca 15.3 were measured at baseline and before each cycle. All adverse events, except peripheral sensory neuropathy, were graded using the National Cancer Institute common toxicity criteria (version 1) at each cycle. Peripheral sensory neuropathy was graded according to the following oxaliplatin-specific scale, modified from Lévi et al:33 grade 1, paresthesias/hypoesthesias of short duration with complete recovery before the next cycle; grade 2, paresthesias/hypoesthesias persisting beyond day 21 without functional impairment; and grade 3, functional impairment persistent between cycles. Tumor evaluation was carried out every three cycles according to standard World Health Organization criteria,34 with the appropriate clinical and the same radiologic examinations, and responses were confirmed within 4 weeks. Responses of 18 patients were reviewed by an external review panel.
In cases of insufficient hematologic function on day 1 of any cycle, treatment was delayed until recovery (neutrophil count ≥ 1,500/μL, platelet count ≥ 100 000/μL). For grade 3 to 4 thrombocytopenia and grade 3 to 4 neutropenia with fever and/or infection, there was a 25% oxaliplatin/20% to 35% 5-FU dose reduction. For grade 3 to 4 gastrointestinal toxicities, only 5-FU was reduced (by 20% to 35%). When grade 2, 3, or 4 hand/foot syndrome occurred, 5-FU was reduced by 20%, 35%, or was discontinued, respectively. In cases of severe neurotoxicity lasting longer than 7 days, oxaliplatin was reduced by 25% then 35%, and if toxicity persisted, the patient was withdrawn from the study. Treatment was continued until disease progression, unacceptable toxicity, patient refusal, or treatment interruption for more than 2 weeks.
RESULTS
Patient Characteristics
A total of 65 patients were enrolled onto the study between October 27, 1997, and November 25, 1999, one of whom did not receive treatment because a brain metastasis was discovered after the patient had been enrolled but before treatment was started. All 64 treated patients were assessed for safety, and 60 patients were assessable for antitumoral response. Patient characteristics are listed in Table 1. The median patient age was 51 years (range, 34 to 71 years), with most patients being in good general condition (96% with performance status of 0 or 1). Eighty-six percent of patients had invasive ductal primary tumors. The median number of organs involved was two (range, one to six organs), with 45 patients (70%) having liver metastases, 30 (47%) having bone metastases, and 22 (34%) having lung metastases.
Patient Characteristics
All patients had received previous treatment with both taxanes and anthracyclines. Thirty-four patients (53%) were taxane-resistant and 22 (34%) were anthracycline-resistant at study entry. Most patients (95%) had undergone previous primary tumor surgery, and 78% had undergone hormonal therapy. The median number of regimens of previous chemotherapy administered was two (range, one to six regimens), 53 patients (83%) having received at least two regimens of previous chemotherapy (including adjuvant). Twenty-two percent of prior chemotherapies administered were given as neoadjuvant therapy and 3% as adjuvant therapy. Eighty-nine percent of patients had undergone previous radiotherapy.
Efficacy
As of May 15, 2001, 60 of the 64 treated patients were assessable. Of the four nonassessable patients, one was found to have a cerebral metastasis in the first treatment cycle, another discontinued after one treatment cycle subsequent to developing an abscess after oxaliplatin extravasation, and two patients withdrew consent for the study after one and two treatment cycles. Objective responses are listed in Table 2. Seventeen partial responses were seen, 26 patients had stable disease (median duration, 5.7 months; range, 2.1 to 24.2 months), and 17 patients had disease progression, giving an objective response rate in the treated population of 27% (95% confidence interval (CI), 16.3% to 39.1%) and 28% in the assessable patient population.
Best Response (WHO criteria)
In the 34 taxane-resistant patients, one patient was not assessable, nine had a partial response, and 14 had stable disease, giving an objective response rate of 26% (95% CI, 12.9% to 44.4%) in this treated population. Among the 22 assessable anthracycline-resistant patients, there were eight documented partial responses and five patients with stable disease, giving an objective response rate of 36% (95% CI, 17.2% to 59.3%) in this treated population. Of the 20 patients who were resistant to both anthracyclines and taxanes, there were eight partial responses and four patients with stable disease, giving an objective response rate of 40% (95% CI, 19.1% to 64.0%) in this patient population.
Of the 17 patients who reported a partial response, 13 (100%) of 13 had a response in liver metastases, three (43%) of seven in the lung, and one (25%) of four in the lymph nodes. When considering the entire treated population, of the 45 patients with liver metastases, 13 (29%) had a partial response, 18 (40%) had stable disease, and 11 (24%) had progressive disease.
In addition, the relationship between tumor response and the extent of decrease in the tumor markers Ca 15.3 and CEA relative to baseline values was examined (Table 3). Analysis showed that the majority of responding patients showed a substantial sustained decrease in both these markers supporting the objective response assessment.
Change in Baseline Ca 15.3 or CEA in Relation to Best Response, by Patient
Time to progression (TTP) was updated until September 2000, whereas overall survival (OS) was updated until January 2001. Median follow-up for survival was 23.1 months (range, 13.6 to 38.5 months). Median TTP in all treated patients (Fig 1) was 4.8 months (95% CI, 3.0 to 6.5), and median OS (Fig 2) was 11.9 months (95% CI, 10.0 to 13.9). In the 17 patients with a partial response, median TTP was 10.2 months (95% CI, 8.5 to 11.9), and median OS was 15.6 months (95% CI, 8.9 to 22.3). Furthermore, these values are not very different when considering both responding patients and those with stable disease (TTP = 7.2 months; OS = 15.0 months). TTP and OS of the anthracycline- or taxane-resistant populations did not differ greatly from those of the whole study population.
Fig 1. TTP.
Fig 2. OS.
Safety
A total of 367 cycles were administered with a median of six cycles administered per patient (range, one to 15 cycles). Of the 64 treated patients, 41 (64%) received at least five cycles, and four patients (6%) received only one cycle (one discontinuing after an oxaliplatin extravasation, one withdrew consent, and two withdrew after early disease progression).
The median relative dose-intensity was 94% for oxaliplatin (range, 64% to 102%) and 92% for 5-FU (range, 17% to 102%). Cycle delays were experienced by 33 (55%) out of the 60 patients who received more than one treatment cycle in a total of 87 (29%) out of 303 cycle intervals, 41 (14%) of which were for more than 7 days. Most cycle delays (62%) were a result of toxicity (usually for hematologic toxicity), and 21% were for personal convenience. Twenty-five patients (42%) of the 60 who received more than one cycle experienced dose reductions in 26 (9%) out of 303 cycles, 11 of whom had dose reductions in both agents.
Seven patients (11%) withdrew as a result of treatment-related adverse events (four for neurotoxicity, one for cardiotoxicity, one persistent grade 3 neutropenia in a progressive disease context, and one patient developed an inguinal abscess after an oxaliplatin extravasation). Four deaths occurred during the study, none of which were treatment related. Of them, three were a result of progressive disease, after having received one, two, or four cycles of study treatment. One woman who received 10 treatment cycles died from a subcortical hematoma after falling and dislocating her shoulder.
Toxicities experienced during treatment are listed in Table 4. The most common toxicity was hematologic, with neutropenia, leukopenia, and thrombocytopenia being prevalent but rarely severe, with grade 3 to 4 being observed in 34%, 16%, and 19% of patients, respectively. The lack of cumulative and/or low morbidity meant that this toxicity had little clinical impact other than dosing adjustments and/or delays. Furthermore, grade 4 toxicity was seldom observed. One patient (2%) experienced a single episode of febrile neutropenia.
Toxicity by Patient
As expected with oxaliplatin-containing combinations, peripheral neuropathy (evaluated according to an oxaliplatin-specific scale) was reported, with grade 2 in 25% of patients, and grade 3 in 8% of patients. The severity of the peripheral neuropathy correlated with the cumulative oxaliplatin dose, with the five patients who experienced grade 3 neurotoxicity receiving a median cumulative oxaliplatin dose of 1,166 mg/m2 (range, 769 to 1,513 mg/m2), which is equivalent to nine treatment cycles.
Gastrointestinal toxicity was reported but was rarely severe, with grade 3 stomatitis, nausea, and diarrhea being seen in 13%, 5%, and 3% of patients, respectively, and grade 3 to 4 vomiting in 3%. Asthenia occurred in 72% of the population, with only 8% of patients experiencing grade 3. Grade 2 alopecia was reported in 34% of patients.
DISCUSSION
This study was motivated by an increasing need for effective alternative second-line therapeutic choices for ABC patients who have failed previous taxane (and implicitly anthracycline) treatment. Both anthracyclines and taxanes are increasingly used in adjuvant/neoadjuvant treatment of patients with locoregional breast cancer, and their combination is now widely used in first-line treatment of ABC. In this multicenter phase II trial, taxane-pretreated patients received 130 mg/m2 of oxaliplatin administered intravenously on day 1 over 2 hours, followed by a 5-FU continuous infusion (1,000 mg/m2/d) from day 1 to day 4 every 3 weeks. It is of note that although the present patient cohort was in good condition at baseline, most patients had several disease sites, a high prevalence of visceral disease, and had been extensively pretreated with systemic chemotherapy and hormonal therapy.
All 64 treated patients had received previous treatment with both anthracyclines and taxanes, and a high proportion were clinically resistant to one or both. Evaluation of the efficacy of this combination showed 17 of the 64 patients achieved a partial response and 26 patients had stable disease, with a median duration of 5.7 months. The objective response rates in all populations analyzed in this study (26% in the whole treated population, 27% in the taxane-resistant population, 36% in the anthracycline-resistant population, and 40% in the concomitantly taxane/anthracycline-resistant population) show that the oxaliplatin/5-FU combination used was active.
This observed activity and its duration are particularly noteworthy given the pattern of disease spread and its extent in the accrued study population, with a high prevalence of both supra and infradiaphragmatic metastatic visceral disease and a relatively low rate of soft tissue disease involvement. It is of note that of the 45 patients with liver metastases at baseline, 13 (28%) had a partial response, and a further 18 (40%) had stable disease as best response. All of the 13 partial response patients with liver metastases showed a response in the liver. The median TTP (4.8 months; 95% CI, 3.0 to 6.5) and median number of administered cycles (six per patient) in this multicenter study are similar to the best reported values in this clinical context.31 The median survival of close to 1 year (11.9 months; 95% CI, 10.0 to 13.9) in breast cancer patients with predominantly visceral disease and who had previously received several regimens considered as the most active in breast cancer is very encouraging. Likewise, the level of activity in the restrictively defined, third party–verified, anthracycline- and taxane-resistant patient cohorts is of note.
The excellent safety profile of the oxaliplatin/5-FU combination was compared in ABC, with a very low incidence of complicated hematologic toxicity. Grade 3 to 4 neutropenia, thrombocytopenia, and leukopenia were seen in 34%, 19%, and 16% of patients, respectively. This hematologic toxicity is equivalent to that observed in other studies of infusional 5-FU (± folinic acid) and oxaliplatin.35 Although hematologic toxicity was of sufficient prevalence to lead to brief dosing delays (29% of cycles were delayed in a total of 55% of patients), dose reductions were minimal. More importantly, morbidity related to such toxicity was almost nonexistent, with only a single reported episode of febrile neutropenia in the 367 cycles administered to heavily pretreated patients.
Of the patients who experienced neurotoxicity, only 8% developed grade 3, despite the fact that some degree of neuropathy was prevalent at accrual (30% of patients) because of prior taxane therapy. This suggests that oxaliplatin-related neurotoxicity was not exacerbated by this regimen. As a consequence, only 11% of patients experienced treatment discontinuations because of adverse events and more than 60% of patients received at least five cycles. The extent of most toxic events falls within the safety range of infusional 5-FU. Furthermore, a major pharmacokinetic interaction between 5-FU and oxaliplatin has been ruled out,36 making this a particularly useful combination therapy.37
This study confirms that the oxaliplatin/5-FU combination reported here is active in this patient population who have exhausted taxane and anthracycline treatment options with a good safety profile compared with previously published regimens.31,38 It is also encouraging that activity was high in patients with liver metastases, which is an acknowledged poor prognostic factor in this patient population. Based on the results presented here, further trials with oxaliplatin and 5-FU in ABC are ongoing, including the evaluation of the activity and safety profile of an infusional 5-FU/oxaliplatin/vinorelbine combination after failure with anthracycline/taxane combinations.
Acknowledgments
Supported by a grant from Sanofi-Synthelabo, Paris, France.
ACKNOWLEDGMENT
We thank Dr Yacine Salhi for statistical analysis.
Footnotes
-
Presented in part at the Thirty-Sixth Annual American Society of Clinical Oncology Meeting, New Orleans, May 20-23, 2000, and the Twenty-Third Annual St Antonio Breast Cancer Symposium, San Antonio, TX, December 6-9, 2000.
- Received June 27, 2001.
- Accepted November 28, 2001.
