AMELIA ISLAND, Fla--Do the available data support the use of high-dose chemotherapy with peripheral blood stem cell support (HDC/PBSC) in advanced breast cancer? That was the question for debate at a session of the Southern Association for Oncology (SAO) meeting.
Providing the case for HDC/PBSC was C. Dean Buckner, MD, of Response Oncology, Inc., Seattle. He presented background data from a large study by Greenberg at M.D. Anderson (J Clin Oncol 14:2197, 1996) showing that approximately 20% of patients with metastatic breast cancer who achieved a complete remission with conventional doxorubicin(Drug information on doxorubicin)-based chemotherapy were disease-free at between 5 and 10 years. "These data document the potential curability of patients with metastatic breast cancer who achieve complete remission," he said, "but with the relatively low complete remission rate in these patients (20% or less), the overall percentage of patients with newly diagnosed metastatic breast cancer who are alive and disease-free at 5 years following conventional chemotherapy is 5% or less."
Dr. Buckner reviewed data from the International Bone Marrow Transplant Registry and studies performed by Response Oncology Inc. These findings suggest that at least 30% to 35% of patients with metastatic breast cancer who received HDC/PBSC while in complete remission would be disease-free at 5 years, an approximately 10% to 15% increase, compared with chemotherapy without HDC/PBSC consolidation.
He also cited data from a published intent-to-treat trial from Response Oncology in which patients agreed to HDC/PBSC at the time of diagnosis. In this trial, approximately one-quarter of 114 patients achieved a complete response to doxorubicin-based induction therapy (AFM), and approximately one-half were in remission after HDC/PBSC, ie, a doubling of the complete remission rate.
"Conversion of patients to complete remission who had failed AFM induction is the best evidence that HDC/PBSC is effective in metastatic breast cancer," Dr. Buckner said. He noted that 15% of patients who were converted to a complete remission with HDC/PBSC were disease free at 4 to 5 years. "None of these patients would be expected to be salvaged with conventional-dose therapies," he said, adding that there were no treatment-related deaths in this trial.
One Randomized Trial
There has been only one randomized phase III trial comparing HDC/PBSC with conventional-dose chemotherapy in patients with newly diagnosed metastatic breast cancer (conducted by Bezwoda and colleagues in South Africa), he said. In this trial, disease-free survival for HDC/PBSC was 20% vs 0% for the control group, with a minimum follow-up of more than 5 years. The control arm had only a 5% complete response rate, which was worse than historical controls.
Dr. Buckner pointed out that there have been no studies of conventional-dose therapies resulting in a 20% 5-year disease-free survival. He also noted that the South African investigators recently reported that patients in this study who had HER2-positive tumors did as well following HDC/PBSC as patients with tumors that were HER2 negative, which was not true for the control group.
"No studies have ever suggested that conventional-dose therapies are superior to HDC/PBSC, and the only arguments against HDC/PBSC therapies relate to cost and the perception of increased toxicities," he said, adding that mortality is 1% or less in Response Oncology breast cancer studies, and costs are decreasing with increasing utilization.
He noted, however, that both conventional-dose chemotherapy and HDC/PBSC are relatively ineffective treatments for patients with newly diagnosed metastatic breast cancer. At most, he said, 20% of these patients will be long-term disease-free survivors. To improve patient outcomes, he urged the development of newer strategies that would improve remission induction, enhance the effectiveness of HDC regimens, or deliver effective treatment after HDC/PBSC.
Dr. Buckner suggested that meaningful comparisons between HDC/PBSC and lower-dose therapies will take large numbers of patients followed for at least 5 years. "It is currently more important to improve overall outcomes of high-risk patients with metastatic breast cancer than to make these comparisons," he said.
Dr. Buckner also discussed the ongoing trials of HDC/PBSC as adjuvant therapy begun in 1989 by Dr. William Peters. The original comparisons suggested a 30% to 40% difference in event-free survival between HDC/PBSC and conventional adjuvant therapy for patients with 10 or more positive axillary nodes.
Several ongoing national and international breast trials are comparing adjuvant HDC/PBSC with adjuvant chemotherapy without PBSC, he said. In all these studies, the intensity of treatment in the control arm has been increased over that administered in the 1980s, which reinforces the general concept that higher-dose therapies are thought to be superior to lower-dose therapies.
"PBSC transplantation is readily available and should be used more widely to evaluate treatment strategies to improve outcomes of patients with breast cancer in the metastatic and adjuvant settings, " Dr. Buckner said. He predicted that, in the future, PBSC will be more widely used to abrogate the hematologic toxicity of HDC and will become a more cost-effective means of supportive care. He suggested that all patients with high-risk breast cancer should have PBSC cryo-preserved at diagnosis.
Why the Enthusiasm?
Frankie Ann Holmes, MD, of Texas Oncology, Houston, subtitled her rebuttal "Why would any (sane) person consider HDC for breast cancer?" She proposed several fanciful reasons for the recent enthusiasm for HDC/PBSC. The first she called the "Mount Everest Principle" (because its there), which may lead people to see things that arent there (ie, the Emperors New Clothes theory).
Some advocates of HDC/PBSC subscribe to the "Star Wars Principle" of using the most aggressive, modern therapy available, she said. And also there is the "Texas Principle" that "Bigger is Better." However, Dr. Holmes said, "we have learned that there is no clinical benefit to higher doses of chemotherapeutic agents; in fact, for cyclophosphamide(Drug information on cyclophosphamide), doses of 600 mg/m² are optimal. Lesser doses lead to lesser antineoplastic efficacy, while higher doses lead to more toxicity."
In addition, she said, inducible resistance to cyclophosphamide can develop after previous treatment with standard cyclophosphamide doses. Finally, combining cyclophosphamide with other alkylators actually may dilute the efficacy of the multiple alkylator regimen.
A perhaps more plausible reason for the enthusiasm for high-dose chemotherapy with transplantation is the dissemination of "promising data" from phase II trials, Dr. Holmes continued. "Phase II trials provide important data about new therapeutic strategies by defining efficacy," she said, "but these trials are, at best, hypothesis-generating, defining the ranges of efficacy." Also, she said, most phase II trials are quite small, and vary in terms of schedules, doses, and patient populations.
"I view the phase II trials of bone marrow transplants with optimistic skepticism," she added, citing unsuspecting bias that eliminated many of the biologically unfavorable tumors in some trials and assured a favorable population in others, resulting in the "cherry picking" of included patients.
Dr. Holmes noted that the early results of the phase III trials, which are comparative trials of high-dose therapy vs no high-dose therapy, "have either not shown any significant improvement in outcome--that is, a 30% or greater increase in disease-free survival--or had too few patients enrolled to make any statistically significant conclusion at all."
A Dutch Randomized Trial
She cited in particular a phase III study from the Netherlands Cancer Institute, Amsterdam, presented at the 1998 American Society of Clinical Oncology (ASCO) meeting (abstract 470). This trial enrolled 97 patients below age 60 who had breast cancer with extensive axillary node metastases. All patients received three courses of upfront conventional chemotherapy consisting of 3-weekly administrations of cyclophosphamide, 500 mg/m², epidoxorubicin, 120 mg/m², and fluorouracil(Drug information on fluorouracil), 500 mg/m².
After surgery, stable and responding patients were randomized to either a fourth course of conventional chemotherapy followed by radiation therapy and 2 years of tamoxifen(Drug information on tamoxifen) (Nolvadex) or identical treatment plus HDC/PBSC after the fourth course of conventional chemotherapy. The high-dose regimen incorporated cyclophosphamide, 6 g/m², thiotepa(Drug information on thiotepa), 480 mg/m², and carboplatin(Drug information on carboplatin) (Paraplatin), 1,600 mg/m².
After surgery, 81 patients were randomized. With a median follow-up of 49 months, there was no significant difference between the high-dose and conventional dose groups with respect to overall survival (79% vs 72% at 4 years) or relapse-free survival (45% vs 56%). The power of the study to detect a 30% difference in relapse-free survival was 80%.
Dr. Holmes noted that newer agents such as taxanes and capecitabine(Drug information on capecitabine) (Xeloda) and new ways to use old and new agents, including modulation of drug schedules, offer incremental strategies that, when combined, may benefit patients.
"Despite overuse of the B (for breakthrough) word, we are standing at the brink of some exciting new therapies for breast and all cancers," Dr. Holmes said. "These are the result of our improved understanding of tumor biology and include, for example, some of the more recently discussed molecules that allow tumors to recruit and retain a blood supply (angiostatin and endostatin) or to penetrate and invade surrounding tissues (the matrix metalloproteinases)."
One interesting aspect of these therapies is that they offer the paradoxical promise of improved or even normal survival but without cure, she said. "These agents may someday allow us to think about cancer as a normal consequence of aging. Imagine treating our tumors the way we treat liver spots!" she said.