High-Dose Chemotherapy for Breast Cancer: Evolving Data

September 1, 1999

Despite a recent decline in incidence and mortality, breast cancer currently develops in one of eight North American women who live to 85 years of age, and remains the major cause of death in American women between the ages of 15 and 54.[1,2]

Despite a recent decline in incidence and mortality, breast cancer currently develops in one of eight North American women who live to 85 years of age, and remains the major cause of death in American women between the ages of 15 and 54.[1,2] Data from historical series and the results of cooperative group adjuvant trials from the 1970s and ’80s indicate that mortality at 10 years in women who have 10 or more involved lymph nodes or a large primary tumor at presentation exceeds 60%. Virtually all patients who initially present with or later develop metastatic breast cancer ultimately die of their disease.[3,4]

In a large series using standard fluorouracil, Adriamycin, and cyclophosphamide (FAC), 17% of women with metastatic breast cancer responded completely, but only 3% were still in remission at 5 years (and only 1.3% at the time of publication).[5] Only complete responders remained free of disease. Thus, complete response, necessary but not sufficient for long-term disease-free survival, is an important surrogate end point. Curative therapy for this group of patients is urgently needed.

Higher chemotherapy doses correlate with response and cure in laboratory models of breast and other cancers.[6] The observation of responses after marrow-ablative chemotherapy doses in patients with disease refractory to conventional chemotherapy led to further clinical trials of high doses in metastatic disease. A recent analysis from the American Blood and Marrow Transplant Registry (ABMTR) showed a 32% progression-free survival rate at 3 years in 640 patients treated in these trials.[7] In addition, pilot data from several centers and the ABMTR[8] showed a 4-year progression-free survival rate ranging from 60% to 65% in patients with high-risk primary breast cancer undergoing transplantation.

Based on these promising data with high-dose therapy (compared with conventional treatment), several centers and cooperative groups began phase III trials comparing high-dose chemotherapy to best available conventional therapy. Of the nine reports of randomized trials of high-dose therapy in patients with high-risk primary or metastatic breast cancer, seven are published only in abstract form, including five trials presented at the American Society of Clinical Oncology (ASCO) meeting in May 1999.

Randomized Adjuvant Trials

The randomized trials of high-dose therapy used in the adjuvant setting are summarized in Table 1.

CALGB Intergroup Study

The Cancer and Leukemia Group B (CALGB) intergroup adjuvant study by Peters et al compared high- vs intermediate-dose CBP (cyclophosphamide, BCNU, and Platinol) after induction therapy with CAF (cyclophosphamide, Adriamycin, and fluorouracil) in women with 10 or more involved lymph nodes and no evidence of metastatic disease (as measured by computed tomographic [CT] scans and bilateral bone marrow aspirates and biopsies).[9] Patients underwent primary excision of their breast tumors and an axillary lymph node dissection prior to the initiation of chemotherapy. Chest wall, supraclavicular, and internal mammary node irradiation followed 6 weeks after chemotherapy. Patients with estrogen-receptor (ER) positive tumors also received hormonal therapy with tamoxifen (Nolvadex) for 5 years.

Among the 884 patients accrued to the trial from 1991 to 1998, the median age of eligible patients was 44 years (range, 22 to 66 years), and the median number of involved lymph nodes was 14. Of the 91 patients who were not randomized, 22 had relapsed and 26 had no insurance coverage. A total of 394 patients were randomized to receive high-dose therapy plus stem-cell transplantation and 391 were assigned to intermediate-dose therapy.

The primary end point of this study is event-free survival. The protocol stipulated an initial analysis after all patients had been followed for at least 3 years. Thus, the data presented at the ASCO meeting are preliminary, and 2 more years of follow-up are needed to assess this primary end point.

Overall, the mortality of transplantation was 7.4% but varied with the experience of the transplant center and increased with patient age. Nonfatal pulmonary and hepatic toxicity of this first-generation BCNU-containing regimen was also substantial. At the median follow-up of 37 months, event-free survival was 68% for the transplant group and 64% for the chemotherapy group.

Discussion of Results—Scientifically, this trial represents a pure comparison between high- and intermediate-dose CBP (since intermediate-dose CBP is not a standard regimen).

The number of relapses was 126 (32%) in the control arm vs 85 (22%) in the transplant group. Patients on both arms of the study had a better progression-free survival than expected for similar historical controls who received conventiona-dose chemotherapy. Eligibility criteria included bilateral marrow biopsies and head CT scans, which are not routinely required for studies of conventional-dose therapies. All patients received consolidative radiotherapy, and patients in both treatment arms received dose-intensive CBP (either at an intermediate dose with growth factor support or at a high dose with stem-cell support).

Survival rates in the two arms were similar at 5 years (71% vs 68% for the transplant and control arms). The lack of a difference in survival is not surprising given the following factors: the expected median time to relapse (months to years) and then to death after relapse (approximately 18 to 24 months); the study’s short median follow-up of 3.6 years; and the potential confounding effect of transplant at relapse. (At the time of study presentation, at least 20 patients on the control arm had already undergone a transplant at relapse.) Because of early mortality, transplant studies often do not initially demonstrate any significant differences in disease-free and overall survival.

Companion QOL Study—A total of 210 consecutive patients in the CALGB intergroup study (about 30% of the study patients) also participated in the companion quality-of-life (QOL) study presented at ASCO 1999. Quality of life was assessed by telephone interviews at baseline and at 3, 12, 24, and 36 months after treatment. The instruments used were the Functional Living Index–Cancer (FLIC), the Symptom Distress Scale, and the Psychosocial Adjustment to Illness Scale (PAIS).

At baseline, QOL scores were similar in women in the two treatment arms. Scores for women in the higher-dose arm were significantly worse on the FLIC and symptom distress scales, and on one subset of the PAIS scale at 3 months, but were equivalent by 1 year and thereafter. Thus, high-dose chemotherapy was associated with worse QOL in the short term only.[10]

South African Study

The South African adjuvant study compared conventional CAF with two high-dose cycles of cyclophosphamide, 4.4 g/m², mitoxantrone (Novantrone), 45 mg/m², and etoposide, 1.5 g/m², plus peripheral blood stem-cell transplantation. This adjuvant trial had no induction phase.[11]

Eligible patients were < 55 years of age, and had a T1 to T3a primary lesion with > 10 involved nodes, a > 5-cm primary tumor with 7 to 9 involved nodes plus one additional poor prognostic factor (such as ER negative disease), or > 7 involved nodes plus two or more first-degree relatives with breast cancer. Eligibility criteria using family history is unusual, but the number of such patients was small, and they were equally distributed among the treatment arms.

Clinical metastases were excluded by bone scan, abdominal ultrasound, and chest x-ray. Chest wall with or without axillary irradiation was given to 11% of the bone marrow transplant (BMT) group and 17% of the CAF group.

Of the 154 patients entered, 1 patient died of toxicity in each arm before day 150. At a median follow-up of 5.3 years, 76 patients had relapsed (21 in the transplant arm and 55 in the control arm; P < .01). The overall survival durations were 400+ weeks for the transplant group and 320 weeks for the control group (P < .01).

Discussion of Results—This unequivocally positive study differs from the other studies in several important ways. First, the duration of follow-up was longer. Second, the experimental arm consisted of two cycles of anthracycline-based high-dose chemotherapy, which differs from the single cycle of alkylator-based therapy used in most other studies. Third, the control arm was conventional CAF rather than the escalated intermediate-dose CBP used in the American intergroup study. Fourth, the patient population had particularly poor prognostic factors, with a low rate of ER and progesterone receptor (PR) positivity. Finally, no conventional-dose induction therapy was administered.

Netherlands Phase II Study

The Netherlands adjuvant phase II study treated 97 women with biopsy-proven apical axillary lymph node involvement with four courses of FEC (fluorouracil, epirubicin, and cyclophosphamide) and then randomized 81 patients to receive either CTCb (cyclophosphamide, thiotepa, and carboplatin) with stem-cell support or an additional cycle of FEC. All of the patients were subsequently treated with surgery, radiation therapy, and 2 years of tamoxifen.[12] However, many of the patients randomized to BMT never actually underwent their assigned therapy.

At a median follow-up of 49 months, most patients in both arms had already relapsed. Although disease-free and overall survival were similar, the small size of the study did not provide sufficient statistical power to rule out differences of less than 30%.[12] This randomized, phase II, feasibility study preceded the launch of a large, recently closed, phase III study, again testing late intensification.

M. D. Anderson Study

The M. D. Anderson Cancer Center randomized 78 patients to eight cycles of FAC with or without two cycles of high-dose cyclophosphamide, etoposide, and cisplatin (Platinol) plus transplantation before the study was closed due to slow patient accrual.[13] Six patients randomized to transplantation received conventional therapy only. Three patients randomized to conventional-dose therapy obtained a transplant elsewhere. No advantage for high-dose chemotherapy has emerged despite a median follow-up that exceeds 78 months, but the study lacks the statistical power to rule out differences of less than 30%.[13]

Scandinavian Trial

Eligibility for the Scandinavian adjuvant trial was based on a “population-based cohort analysis” showing a > 70% chance of relapse, which encompassed patients with: (1) more than eight positive lymph nodes; (2) more than five positive nodes with ER negative disease and a high S-phase fraction; (3) a positive bone scan with negative plain films; or (4) microscopic bone marrow involvement. Thus, the study included some patients with early stage IV disease.

A total of 525 patients received three cycles of induction FEC followed by randomization to one high-dose cycle of CTCb vs nine cycles of dose-intensive “tailored” FEC; in the latter group, doses of epirubicin and cyclophosphamide were individualized according to nadir blood counts (based on the known significant interpatient variability in serum drug levels with fixed chemotherapy doses). Doses of epirubicin were escalated to 120 mg/m², and doses of cyclophosphamide were increased to 1,800 mg/m², with a fixed dose of fluorouracil of 600 mg/m² per cycle. The transplant group received 600, 60, and 600 mg/m²of fluorouracil, epirubicin, and cyclophosphamide, respectively, for the first two cycles, with stem cells collected after a third cycle of 600, 60, and 1,200 mg/m², respectively, of the three drugs. Following the completion of chemotherapy, both groups received local radiotherapy and, for those with ER positive disease, tamoxifen.[14]

The planned cumulative doses for the tailored therapy arm actually exceeded those for the BMT arm. As a result, this study compared one cycle of high-dose therapy to six cycles of intensified, intermediate-dose chemotherapy with a higher cumulative chemotherapy dose. Thus, a superior disease-free survival and overall survival for patients who received tailored therapy would support the importance of cumulative dose over early peak dose.

At a median follow-up of only 2 years, eight patients (3%) in the tailored-dose arm have developed leukemia or myelodysplasia. Thus, additional cases are likely. In the BMT arm, there have been two treatment-related deaths but no leukemias. Stem cells collected after three cycles of chemotherapy may be less damaged than those exposed to nine relatively high-dose cycles.

A total of 66 patients in the tailoredchemotherapy arm and 92 in the transplant group have relapsed. Survival in the two groups is equivalent, at approximately 65%. The early follow-up of this trial precludes firm conclusions from being made. Moreover, any conclusions that might be reached could not address the relative effectiveness of transplant vs conventional therapy.

Conclusions Regarding the Adjuvant Trials

Of the three mature studies, the South African study shows significant benefits of high-dose therapy with respect to disease-free and overall survival. Two smaller studies show no advantage for high-dose therapy but cannot exclude differences of up to 30%.

The preliminary data from the CALGB intergroup trial show interesting differences in disease-free survival, the primary end point of the study. The Swedish study compares two different high-dose regimens and, therefore, cannot address the question of high- vs conventional-dose therapy.

Randomized Trials in Metastic Disease

Results of the four randomized trials of high-dose chemotherapy in patients with metastatic breast cancer are summarized in Table 1.

Philadelphia Intergroup Study

In the Philadelphia intergroup study, patients received four to six cycles of either CAF (for patients with no prior anthracycline therapy) or CMF (cyclophosphamide, methotrexate, and fluorouracil). Patients who were in complete or partial remission after initial therapy were randomized to either late intensification using cyclophosphamide, 6,000 mg/m², thiotepa, 500 mg/m², and carboplatin, 800 mg/m² (CTCb regimen) plus stem-cell transplantation, or maintenance chemotherapy with CMF for 24 months or until disease progressed.[15] Patients with bone-only disease were eligible for participation if they had stable radiographic disease and improved symptoms.

Of the 296 patients eligible for randomization, 199 were randomized (36% of the 535 patients entered) and 82 refused to receive their randomized therapy. Of the 199 randomized patients, 110 were assigned to transplantation and 89, to CMF chemotherapy. An additional 18% of the randomized patients were ineligible or did not receive their assigned treatment. (Of these patients, 4 refused to undergo stem-cell transplantation and 11, CMF.) Of the 101 patients who underwent transplantation, 1 died of regimen-related toxicity.

Of the eligible patients, 11% achieved a complete response after initial chemotherapy for metastatic disease and 47% attained a partial response. No significant differences between autologous bone marrow tranplantation and CMF were seen with respect to 3-year progression-free survival rates (6% vs 12%) or overall survival rates (32% vs 38%).

Discussion of Results—Of the 55 patients in complete response, only 45 were actually randomized. Although the Philadelphia study showed no advantage for patients in complete response who received high-dose therapy, the Duke study, which randomized a much larger number of patients in complete response, did demonstrate a significant disease-free survival benefit of high-dose therapy.

Given the less than 1% mortality, similar toxicity, and identical survival rates in the two arms of the Philadelphia intergroup study, many patients might prefer a short, intense treatment regimen of up to 2 years of monthly chemotherapy. A QOL component of this study is underway.

The substantial dropout rate in this study—with comparisons on only 184 patients analyzed out of 535 patients entered—appears to have resulted in a treatment group with poor prognostic features for whom high-dose therapy is already known to result in early relapse.[7]

South African Trial

The South African trial randomized 90 patients with metastatic breast cancer regimen directly to two cycles of a high-dose anthracycline-based regimen vs conventional-dose therapy; no induction therapy was given. Most of the patients had poor prognostic features.

Patients randomized to the high-dose therapy arm had superior complete response rates, as compared with those assigned to conventional therapy (51% vs 4%). Disease-free survival time also was better in the high-dose therapy group (80 vs 34 weeks), as was overall survival time (90 vs 45 weeks).[16] In a more recent follow-up of the trial, nine patients randomized to the high-dose arm remained in continuous complete remission for more than 5 years.[16,17]

A higher proportion of patients in the high-dose arm received tamoxifen, which was prescribed only for complete responders. This may have biased the results of the study.

Duke Crossover Study

In a Duke crossover study, 120 of 453 patients with metastatic breast cancer treated with standard-dose AFM (Adriamycin, fluorouracil, and methotrexate) attained a complete response. Of patients who had a complete response, 98 were randomized to either immediate high-dose therapy or high-dose therapy at the time of relapse. Disease-free survival was significantly improved in the immediate BMT group, but survival was longer in the group receiving delayed transplantation.[18]

French Study

The French metastatic study, PEGASE 04, enrolled 61 patients who had responded to four to six cycles of their first conventional-dose chemotherapy for metastatic breast cancer. Of these 61 patients, 32 were randomized to 45 mg/m² of mitoxantrone, 120 mg/kg of cyclophosphamide, and 140 mg/m² of melphalan (Alkeran), and 29 were randomized to continued conventional-dose chemotherapy.[19]

Median progression-free survival durations were 20 and 35.3 months in the standard- and intensive-dose groups, respectively (P = .06). The relapse rates in the two groups were 79% vs 51% at 3 years and 91% vs 91% at 5 years, respectively. Median overall survival times in the standard- and intensive-dose groups were 20 and 43 months, respectively, with overall survival rates of 18% and 30% at 5 years (P = .12).

Conclusions Regarding the Metastatic Disease Studies

All of these studies are small but mature. The South African study demonstrates significantly positive results of high-dose chemotherapy with regard to disease-free and overall survival. The Duke and French studies show significant benefits with respect to disease-free survival, and the French study shows a nonsignificant doubling of median survival. Survival in the Duke study (because of the crossover design) does not address the question of high- vs conventional-dose therapy. The Duke trial does, however, address the issue of disease-free survival in patients in complete response, with more than double the number of complete responders than in the Philadelphia trial (98 vs 45).

In contrast, the Philadelphia trial differs from the other three trials in that it is negative for both disease-free and overall survival, perhaps because of the very high fraction of patients who dropped out or because of the design of the study, which compared 1 cycle of high-dose therapy with up to 24 cycles of conventional therapy.

Overall Conclusions

Treatment-related mortality for both the high-dose and control groups in the randomized trials is consistently 0% to 2.5 % (compared to a 22% in the 1989 ABMTR data[7]); the only exception to this is the CALGB intergroup study, which reports a treatment-related mortality of 7.4%. For 1993 through 1995, the ABMTR reports treatment-related mortality of 3% to 5%.[7] In several studies, QOL declines at the time of transplant ation but returns to normal about 3 months after completion of therapy.

Because of the very different designs of the adjuvant trials and short follow-up of many of these trials, any conclusions remain controversial. The two smallest adjuvant studies (Dutch and M. D. Anderson studies), which compared standard FAC or FEC with or without high-dose chemotherapy, reported no differences in survival.

The two largest studies, the US study comparing high-dose with intermediate-dose therapy, and the Scandinavian study comparing one high-dose cycle with six intermediate-dose cycles, have short follow-up periods of just 2 and 3.6 years, respectively. Furthermore, the Scandinavian study does not even address the primary question of high- vs standard-dose chemotherapy.

The two studies comparing four to six cycles of induction chemotherapy followed by one cycle of high-dose therapy vs nine cycles (Scandinavian study) or up to 24 cycles (Philadelphia study) of standard treatment have identical survival curves. These results suggest that brief, intense chemotherapy may be equivalent but not superior to prolonged conventional-dose chemotherapy.

The South African trials in high-risk and metastatic disease, which used two high-dose cycles of an anthracycline-based regimen vs conventional-dose CAF chemotherapy, are unequivocally positive. However, these trials differ from the others with respect to design (eg, the use of two cycles of a high-dose anthracycline-based regimen, the lack of an induction regimen), and included patients with poor prognostic features.

Given the endogenous and acquired drug resistance of breast cancer, a randomized trial of induction vs none may be appropriate. No current randomized trial formally tests this hypothesis.

Additional follow-up of these nine randomized trials, and particularly the completion and reporting of other ongoing, large, randomized, international trials will provide more reliable information to determine the role of high-dose chemotherapy in the management of high-risk primary and metastatic breast cancer.


1. Landis SH, Murray T, Bolden S, et al: Cancer statistics, 1999. CA Cancer J Clin 49:8-11, 1999.

2. Wingo PA, Reies LAG, Rosenberg HM, et al: Cancer incidence and mortality, 1973-1995: A report card for the United States. Cancer 82:1197-1207 1998.

3. Clark G, Sledge GW, Osborne CK, et al: Survival from first recurrence: Relative importance of prognostic factors in 1,015 breast cancer patients. J Clin Oncol 5:55-61, 1987.

4. Mick R, Begg CB, Antman K, et al: Diverse prognosis in metastatic breast cancer: Who should be offered alternative initial therapies? Breast Cancer Res Treat 13:33-38, 1989.

5. Greenberg P, Hortobagyi G, Smith T, et al: Long-term follow-up of patients with complete remission following combination chemotherapy for metastatic breast cancer. J Clin Oncol 14:2197-2205, 1996.

6. Frei III E, Canellos GP: Dose, a critical factor in cancer chemotherapy. Am J Med 69:585-594, 1980.

7. Antman K, Rowlings P, Vaughn W, et al: High-dose chemotherapy with autologous hematopoietic stem cell support for breast cancer in North America. J Clin Oncol 15:1870-1879, 1997.

8. Rowlings P, Antman K, Fay J, et al: Prognostic factors for outcome of autotransplants in women with high-risk primary breast cancer (abstract). Proc Am Soc Clin Oncol 16:117a, 1997.

9. Peters WP, Rosner G, Vredenburgh J, et al: A prospective, randomized comparison of two doses of combination alkyating agents as consolidation after CAF in high-risk primary breast cancer involving ten or more axillary lymph nodes: Preliminary results of CALGB 9082/SWOG 9114/NCIC MA-13 (abstract). Proc Am Soc Clin Oncol 18:1a, 1999.

10. Winer EP, Herndon J, Peters WP, et al: Quality of life in patients with breast cancer randomized to high-dose chemotherapy with bone marrow support vs intermediate-dose chemotherapy: CALGB 9066 (companion protocol toCALGB 9082) (abstract). Proc Am Soc Clin Oncol 18:412a, 1999.

11. Bezwoda WR: Randomized, controlled trial of high-dose chemotherapy vs standard dose chemotherapy for high-risk, surgically treated, primary breast cancer (abstract). Proc Am Soc Clin Oncol 18:2a, 1999.

12. Rodenhuis S, Richel KJ, van der Wall E, et al: Randomized trial of high-dose chemotherapy and hematopoietic progenitor-cell support in operable breast cancer with extensive axillary lymph-node involvement. Lancet 352:515-521, 1998.

13. Hortobagyi GN, Buzdar AU, Champlin R, et al: Lack of efficacy of adjuvant high-dose tandem combination chemotherapy for high-risk primary breast cancer: A randomized trial (abstract). Proc Am Soc Clin Oncol 17:12, 1998.

14. Bergh J: Results from a randomized adjuvant breast cancer study with high-dose chemotherapy with CTCb supported by autologous bone marrow stem cells vs dose escalated and tailored FEC therapy (abstract). Proc Am Soc Clin Oncol 18:2a, 1999.

15. Stadtmauer EA, O’Neill A, Goldstein LJ, et al: Phase III randomized trial of high-dose chemotherapy and stem cell support shows no difference in overall survival or severe toxicity compared to maintenance chemotherapy with cyclophosphamide, methotrexate, and 5-fluorouracil for women with metastatic breast cancer who are responding to conventional induction chemotherapy: The Philadelphia intergroup study (abstract). Proc Am Soc Clin Oncol 18:1a, 1999.

16. Bezwoda W, Seymour L, Dansey R: High-dose chemotherapy with hematopoietic rescue as primary treatment for metastatic breast cancer: A randomized trial. J Clin Oncol 13:2483-2489, 1995.

17. Bezwoda WR: Primary high-dose chemotherapy for metastatic breast cancer: Update and analysis of prognostic factors (abstract). Proc Am Soc Clin Oncol 17:115a, 1998.

18. Peters W, Jones R, Vredenburgh J, et al: A large prospective randomized trial of high-dose combination alkylating agents (CPB) with autologous cellular support as consolidation for patients with metastatic breast cancer achieving complete remission after intensive doxorubicin-based induction therapy (AFM) (abstract). Proc Am Soc Clin Oncol 15:121, 1996.

19. Lotz JP, Cure H, Janvier M, et al: High-dose chemotherapy with hematopoietic stem cells transplantation for metastatic breast cancer: Results of the French protocol Pegase 04 (abstract). Proc Am Soc Clin Oncol 18:43a, 1999.