Reduction in new metastases in breast cancer with adjuvant clodronate treatment 

Authors Diel I, Solomayer E, Costa S, Gollan C, Goerner R, Wallwiener D, Kaufmann M, Bastert G. 
Source New England Journal of Medicine. 339:357-63. August 6, 1998. 
Institutions University of Heidelberg and University of Frankfurt, Germany. 
Support Boehringer Mannheim. 


Bisphosphonates inhibit osteoclast-mediated bone resorption. In patients with breast cancer metastatic to bone they have been shown to reduce the incidence of hypercalcemia, bone pain and fractures. One bisphosphonate, clodronate, also reduced the number of new skeletal metastases in patients with advanced breast cancer but no preexisting bone metastases. 

This study was designed to investigate the effect of clodronate on bony and visceral metastases in patients with primary breast cancer who had no evidence of visceral or bone metastases but who did have tumor cells in a bone marrow biopsy specimen obtained at the time of enrollment.



Inclusion criteria:

  • Primary breast cancer of any size (T0-T4), with or without ipsilateral lymph nodes (N0-N2).
  • Immunocytochemical evidence of tumor cells in a bone marrow aspirate (presence of tumor-associated glycoprotein TAG12).
Exclusion criteria:
  • Metastatic disease (M1) or ipsilateral internal mammary lymph nodes (N3).
  • Other malignancies, chemotherapy, skeletal disease, severe hepatic or renal disease and pregnancy.

All patients underwent either mastectomy or breast-conserving surgery with radiation, as well as axillary lymphadenectomy and bone-marrow biopsy. 

Patients were then randomized to receive or not to receive treatment with clodronate, 1600 mg daily for two years (this was not a placebo controlled study). 

After randomization, and based on prognostic factors (lymph node status, menopausal status, estrogen receptor status and tumor size), the decision was made whether or not to administer adjuvant therapy and what specific therapy to administer. Decisions about adjuvant therapy were not modified by the results of the bone marrow biopsy. 

All patients who subsequently developed bone metastases were treated with clodronate. Patients with disease progression / metastases were treated with anti-neoplastic agents and/or hormonal therapy. 


Patients were seen every 3-4 months for history, physical examination and bloodwork and yearly for bone scanning, chest x-ray, liver ultrasonography and mammograms. 

The primary study endpoints were the incidence and number new bony and visceral metastases and the time to their appearance.



302 patients were randomized, 157 to the clodronate group, 145 to the control group. Patients excluded from evaluation (non-compliance, drug side effects, loss of follow up) were 3 in the control group and 15 in the clodronate group. 

Prognostic factors were well matched between the groups. 

The median age of patients was 51 years. In the control group 39% were pre-menopausal vs. 36% in the clodronate group. 

Histologic grade was I or II in 73% of the control group vs. 68% of the clodronate group. 

Tumors were stage T1 or T2 in 83% of patients, lymph node status was negative in 47%, estrogen receptor status was positive in 73%, progesterone status was positive in 63%. 

19% of patients received no adjuvant chemotherapy. The remaining 81% received one of five regimens, mainly tamoxifen and/or cytoxan-methotrexate-fluorouracil, well balanced between the two groups. 


Median follow-up was 36 months. 

Bone metastases: 12 (8%) in the clodronate group, vs. 25 (17%) in the control group (p=0.003). Mean number of bony metastases per patient: 3.1 vs 6.3 (p=0.004). 

Visceral metastases: 13 (8%) vs 27 (19%) (p=0.003). 

Any metastasis (some patients had both): 21 (13%) vs 42 (29%) (p<0.001). 

Overall survival

There were 6 deaths in the clodronate group (4%) vs 22 in the control group (15%) (p=0.001). 

These results were confirmed in Kaplan-Meier curves of survival without metastases and overall survival. 


Author's discussion

The authors make a number of points in their discussion: 
  • These women all had a high risk of distant metastasis, based on the presence of tumor cells in the bone marrow. They state that tumor cells were found in the marrow in 55% of node-positive patients and in 31% of node negative patients. 

  • The beneficial effect of bisphosphonates on bony metastases has been suggested by animal experiments and by trials in women with documented bone metastases.  In the early phase of metastasis to bone, tumor cells activate osteoclasts. There is evidence that growth factors which may promote the growth of tumor cells are released when bone is degraded. Bisphosphonates inhibit the activity of osteoclasts, which may account for some of the favorable effects seen here.  Other possible mechanisms include the induction of apoptosis and alteration of adhesion molecules on tumor and bone cells.

  • Because of the limited follow-up period, it is unclear whether the reduction in bone metastases seen was due to a delay in their appearance or an actual prevention of their development.

  • Unlike the effect on bone metastasis, the effect on visceral metastasis was unexpected. Possible explanations include the same mechanisms that are thought to apply to bone metastasis (apoptosis and alteration of adhesion molecules), as well as some sort of synergy with cytotoxic agents.


In this study, clodronate was administered orally for two years to patients with breast cancer, without evidence of metastatic disease but with tumor cells found on bone marrow biopsy. During a mean follow-up of 36 months, there was a significant decrease in the incidence of new bony metastases and an increase in the time to appearance of this complication. Surprisingly, a similar effect was noted on visceral metastases and on overall survival. 

Complications from breast cancer can occur many years later. Thus, as the authors point out, it is impossible to tell whether their results indicate definitive prevention of metastasis and death for some women, or a delay in the occurrence of these events.  Looking at the shape of the Kaplan-Meier curves in the article, one has the impression that the divergence is due to a delay in the occurrence of events. However, even if cure is preferable to delay, a significant delay is much better than no delay. 

It is surprising that this trial was randomized but not blinded or placebo-controlled. It is not stated that the radiologists who interpreted the bone scans were blinded to the treatment assignment. This in troduces some concern about observer bias in interepretation of results and symptoms, when patients came for follow-up. 

There is another aspect of the study design that I find troubling. In order to participate, women had to undergo an iliac-crest bone-marrow aspirate. It is specifically stated that adjuvant chemotherapy decisions did not take into account the results of the bone-marrow biopsy. Thus, it would seem that those women who were randomized to the control group received no benefit from this procedure and the information it provided.  This is particularly troubling since the presence of tumor cells in the bone marrow indicates a worse prognosis and thus might justify a more aggressive adjuvant therapy approach. Despite the presence of tumor cells in the marrow, 19% of the women received no adjuvant therapy at all, including tamoxifen.  Would more aggressive adjuvant therapy in all patients have diminished the magnitude of clodronate's effect seen in this study? Unfortunately, the authors do not give any data on the effect of clodronate therapy in the various subgroups of adjuvant therapy. 

Finally, since all patients in this study had documented tumor cells in the bone marrow, it is not clear how the results would apply to large proportion of women (up to 70% of those with negative axillary nodes) who do not have tumor cells in their marrow. 

These objections aside, the role of bisphosphonates in the treatment of breast cancer is likely to increase, with many questions that remain to be answered concerning the effectiveness of various agents, their combination with other therapies and the appropriate target populations. 

September 4, 1998 


References related to this article from the NLM's PubMed database. 

Reader Comments

December 28, 1998

Letters to the editor about this article, from the New England Journal of Medicine.

Submit a comment

Journal Club homepage

Site Meter