Combination Therapy
Doxorubicin/Cisplatin
Several studies have shown impressive antitumor activity of combinations of paclitaxel(Drug information on paclitaxel) with doxorubicin(Drug information on doxorubicin)[40-45] and cisplatin(Drug information on cisplatin) (Platinol)[46,47] against metastatic breast cancer and have begun to elucidate the potential toxicities of these combinations. It has become apparent in these trials that dose, duration of infusion, and sequence of administration are all important determinants of toxicity. Randomized trials, such as the intergroup study of paclitaxel vs doxorubicin vs the combination of the two agents with G-CSF support, are critical to gauge the relative value of such combinations over single-agent paclitaxel.
Edatrexate
Edatrexate is an analog of methotrexate(Drug information on methotrexate) that competes for the folate-binding site of the enzyme dihydrofolate reductase, and thus, indirectly blocks the synthesis of nucleotides. It possesses potential preclinical advantages over methotrexate in that it demonstrates greater selective entry and intracellular conversion to polyglutamate forms in neoplastic cells compared to other antifolates.[15] In addition, edatrexate has shown promising single-agent activity against metastatic breast cancer in previous clinical trials.[15-17]
In vitro data from our center have demonstrated that the sequence of edatrexate followed by paclitaxel showed marked synergism in inhibiting the growth of SKBR-3 human breast adenocarcinoma cells,[18,19] while the reverse schedule showed antagonism. We therefore evaluated the sequential combination of edatrexate and paclitaxel in a phase I-II clinical trial in patients with stage IV breast cancer.[48] Edatrexate doses of up to 350 mg/m² were well tolerated in combination with paclitaxel at 175 mg/m² via a 3-hour infusion without hematopoietic growth factor support, with both agents recycled every 21 days. Preliminary data show eight responses (three complete, five partial) among the first 12 evaluable patients (edatrexate dose range, 180 to 270 mg/m²).
Cyclophosphamide
For chemoresponsive metastatic breast cancer, single-cycle, conventional high-dose chemotherapy regimens requiring autologous stem-cell support have produced complete responses in up to 50% of patients, and yet the majority of these patients ultimately develop recurrent disease. Motivated by the apparent failure of a single high-dose application of chemotherapy to eradicate all viable malignant cells in prior clinical trials, we applied the concepts of the Norton-Simon hypothesis and the Gompertzian model of breast cancer kinetics[49] in a series of studies. We evaluated the delivery of multiple courses of high-dose alkylating agents at short intertreatment intervals in patients with responsive metastatic breast cancer.[50]
Notably, our group has demonstrated that the addition of paclitaxel (250 mg/m²) to high-dose cyclophosphamide(Drug information on cyclophosphamide) (3 g/m²) does not compromise the mobilization of CD34+ peripheral blood progenitor cells (median, 16.22 E6/kg/leukapheresis), as compared with 3 g/m² of cyclophosphamide alone (2.64 E6/kg/leukapheresis).[51] Hence, we are presently evaluating the incorporation of paclitaxel into a high-dose sequential regimen in chemosensitive metastatic breast cancer. This regimen consists of tandem cycles of high-dose cyclophosphamide plus paclitaxel followed by tandem cycles of high-dose thiotepa(Drug information on thiotepa) plus paclitaxel with peripheral blood progenitor cells for hematologic rescue (MSKCC IRB protocol 94-77).
Monoclonal Antibodies
There has recently been an expansion of knowledge about the role that certain oncogenes, growth factors, and growth factor receptors play in breast cancer.[52] One of the best-studied growth factor receptor systems in breast cancer includes the epidermal growth factor receptor (EGFR) and the closely related HER-2/neu receptor, both of which possess intrinsic tyrosine kinase activity. When directed against both these receptors, MoAbs inhibit the growth of breast cancer cells overexpressing the target receptor.[52,53] Over the past several years, compelling experimental data have suggested that combining certain chemotherapeutic agents with MoAb-mediated blockade of either EGFR or HER-2/neu receptors can eradicate well-established human tumor xenografts resistant to either treatment given singly.[20,21,54]
Significant antineoplastic effects are observed when human breast cancer xenografts are exposed to paclitaxel in combination with either anti-EGFR or anti-HER-2/neu MoAb.[21] This strong synergy is achieved with no increased toxicity in the animal model. Although the mechanisms for the apparent supra-additive effects of certain chemotherapeutic agents and MoAbs are still being investigated,[55] these data provide a lead for translation into the clinic. Recent data indicate that downstream events in the signal transduction pathway may be involved in paclitaxel-induced apoptotic cell death.[56] We are presently evaluating the human-murine chimeric anti-EGFR MoAb C225 in combination with paclitaxel in a phase I trial in patients with metastatic breast cancer.
Growth Factor Receptors and Clinical Paclitaxel Sensitivity
There is a growing body of predominantly preclinical information suggesting a relationship between growth factor receptor expression and chemosensitivity or resistance.[57] Among patients receiving higher doses of anthracycline-based adjuvant therapy, those with primary breast cancers overexpressing the HER-2/neu receptor had improved disease-free and overall survival compared with those that lacked overexpression.[58] To better define this relationship for paclitaxel, we conducted a multivariate analysis of immunohistochem- ical HER-2/neu expression and clinical taxane sensitivity in patients receiving single-agent paclitaxel (and docetaxel(Drug information on docetaxel)) in clinical trials over the last 5 years at MSKCC.[59]
Preliminary analysis showed that 51 (40.5%) of 126 patients treated with either single-agent paclitaxel or docetaxel had tumor overexpression of HER-2. The overall response rate for all patients was 46.8% (59/126). The response proportion was 58.8% (30/51) if HER-2 was overexpressed, compared with 38.7% (29/75) if HER-2 was negative (Mantel-Haenszel test, P = .027).
Among the factors assessed in conjunction with tumor response, visceral-dominant disease, low KPS, and extensive prior therapy correlated with a poor clinical response. Among these, HER-2 overexpression was positively correlated with a low KPS, and a low KPS with extensive prior therapy. It appears, therefore that HER-2 overexpression in metastatic breast cancer seems to confer sensitivity rather than resistance to taxanes, despite a positive correlation of HER-2 positivity with poor prognostic features. We are presently expanding our analysis to address other relevant oncogene modulators of paclitaxel cytotoxicity, including regulators of apoptotic cell death.[56,60]
Adjuvant Therapy of Early-Stage Breast Cancer
The significant activity and safety of paclitaxel noted among patients with advanced disease has motivated us to incorporate the agent into a postoperative adjuvant chemotherapy regimen. We sequence the scheduling of active therapeutic components in the adjuvant setting, as suggested by mathematical models of tumor kinetics[49] and substantiated by a clinical trial.[61] We have previously demonstrated the feasibility of sequential administration of doxorubicin and high-dose cyclophosphamide as adjuvant therapy for patients with resectable stage II-III breast cancer with four or more involved axillary lymph nodes.[62] The recurrence-free survival curve noted thus far is encouraging; with a median follow-up time of 895 days, a 65% recurrence-free survival rate has been noted among 60 patients with a median of nine involved axillary nodes.
This experience, coupled with paclitaxel's activity and partial non-cross-resistance with doxorubicin, led us to incorporate it into an adjuvant chemotherapy regimen for women of the same risk category.[23] Forty-two patients with 4 or more positive axillary lymph nodes (median, 8 nodes; range, 4 to 25) have received the regimen of rapidly sequenced doxorubicin (90 mg/m²), paclitaxel (250 mg/m², 24-hour schedule), and high-dose cyclophosphamide (3,000 mg/m²), all administered with G-CSF support, as shown in Figure 1.[22] The median delivered dose intensity for each component of this regimen has been 100% of planned intensity.
Approximately two-thirds of patients had to be hospitalized at least once during the regimen, most commonly for febrile neutropenia. A similar proportion of patients required red blood cell transfusion, and 10% required platelet transfusion. The more frequent grade 3 nonhematologic toxicities included fatigue (24%), bone pain (24%), stomatitis (17%), dermatologic reactions (17%) neurosensory effects (15%), nausea (12%), and diarrhea (7%). Serial gated radionuclide heart scans showed no decline in cardiac ejection fraction, and no clinical cardiotoxicity was noted.
At a median follow-up of 448 days from surgery (range, 82 to 632 days), 7% of patients have relapsed (Figure 2). In an effort to optimally integrate paclitaxel into adjuvant systemic therapy, we performed a study randomizing patients to receive either a slightly modified version of the above sequential regimen or single-agent doxorubicin ×3 followed by concomitant paclitaxel and high-dose cyclophosphamide.[23] The median number of positive lymph nodes was 8 (range, 1 to 35), and the median tumor size was 2.3 cm (range, 0.4 to 8.0 cm). All patients received G-CSF. A total of 41 patients were enrolled; 21 were randomized to receive sequential paclitaxel and cyclophosphamide (arm A) after dose-dense doxorubicin, and 20 patients received paclitaxel and cyclophosphamide concomitantly (arm B).
At a median follow-up of 10 months (range, 6 to 15 months), there have been no relapses, deaths, or instances of cardiac toxicity. The concomitant administration of paclitaxel and cyclophosphamide was associated with greater toxicity, with 16 of 20 patients (89%) hospitalized for toxicity compared with 4 (19%) of 21 patients on the sequential arm. The concomitant regimen required greater dose-reduction and delay after doxorubicin, while offering no known advantage over sequential administration.
Therefore, the appropriate regimen for phase III testing is sequential doxorubicin, paclitaxel, and cyclophosphamide, and a trial testing this regimen is ongoing (Figure 3). This trial compares sequential dose-dense chemotherapy with doxorubicin, paclitaxel, and high-dose cyclophosphamide with G-CSF support to a more standard doxorubicin/cyclophosphamide doublet followed by high-dose chemotherapy (STAMP I or V) requiring peripheral blood progenitor cell (and G-CSF) support.
This study should complement the information being obtained from the important intergroup randomized trial (Figure 4), which is evaluating the value of four cycles of paclitaxel, administered via a 3-hour infusion, after delivery of one of three dose levels of doxorubicin/cyclophosphamide (60/600, 75/600, or 90/600 mg/m², respectively) for four courses as adjuvant chemotherapy for node-positive early-stage breast cancer. It is hoped that these results will be further augmented by those of the NSABP B-28 trial, which is using four courses of paclitaxel or no further therapy after four courses of standard doxorubicin/cyclophosphamide adjuvant therapy.
Conclusions
By the end of the 1990s, a number of large multicenter trials will have provided the answers to important questions regarding the optimal application of single-agent paclitaxel (dose and schedule), its role in relation to other active agents and regimens, its comparative impact on QOL (ie, therapeutic index), and the potential benefits of combination regimens. Importantly, the potential contribution of paclitaxel to improving our ability to cure early-stage breast cancer will begin to be appreciated when survival curves from ongoing and soon-to-be-completed adjuvant trials mature.
Studies at MSKCC and elsewhere are attempting to characterize in vivo resistance mechanisms to taxanes--efforts that may result in mechanism-directed strategies to overcome resistance and/or to guide the development of analogs. Translational research involving growth factors and their receptors promises to exploit an expanding knowledge of autocrine and paracrine pathways. In both the laboratory and the clinic, there are numerous reasons for optimism regarding the future contribution of paclitaxel to the fight against breast cancer.
