Patient-Individualized Chemotherapy in Breast Cancer

Institution: University of California, Los Angeles
Investigator(s): Daniel Silverman, M.D., Ph.D. -
Award Cycle: 2001 (Cycle VII) Grant #: 7KB-0155 Award: $296,994
Award Type: New Investigator Awards
Research Priorities
Detection, Prognosis and Treatment>Innovative Treatment Modalities: search for a cure



Initial Award Abstract (2001)
The most frustrating part of treating breast cancer for both the patient and clinician is that many therapies fail to work as planned. This is often the result of the tumor being non-responsive to the chosen drug, as there is currently no way to consistently predict responsiveness prior to initiating treatment. Selection of breast cancer chemotherapy agents, now based on a few general parameters (tumor histology, clinical stage, receptor and antigenic status) leads to some patients receiving toxic drugs from which they fail to benefit, and possibly experiencing delay in getting a regimen from which they would benefit more. Thus, better information on predictors of tumor response to specific chemotherapy compounds in individual patients is desperately needed.

This project will use an advanced method of tumor imaging, called PET (positron emission tomography), to track chemotherapy agents labeled with radioactive fluorine (18F). The fluorine tag will allow us to accurately measure the concentration and time course of each chemotherapy agent in tumors and normal tissues. These experiments will be first performed in mice that have been implanted with small human breast tumors, and will take advantage of a special technique developed at UCLA, called microPET imaging. We hypothesize that the very small (non-toxic) doses of each labeled chemotherapy agent used in microPET imaging of each tumor will predict the subsequent degree of tumor shrinkage and elimination.

Two chemotherapy compounds we will focus on are cyclophosphamide and paclitaxel. The first challenge will be to show that covalent incorporation of 18F into the drug can be accomplished without significantly altering its biodistribution properties. The next challenge will be to demonstrate that the distribution of the tracer doses of the labeled compounds can serve to accurately forecast how the usual toxic doses will distribute into the tumors and the rest of the body. Finally, we will test the ability of our distribution measurements to improve the accuracy with which the effectiveness of each chemotherapy agent is predicted.

We anticipate that performing these planning scans in patients about to undergo chemotherapy will have the potential to alleviate much suffering in cancer patients, and lead to greater clinical success in treating the disease. Using PET to measure potential therapeutic agents in tracer amounts will allow a clinician to survey a battery of drugs, then select the compound and the dose that is individualized for each woman to most effectively combat her disease. Our goal is to first demonstrate this method in mice that are bearing human tumors, and then extend the approach into human clinical trials.


Final Report (2004)
Introduction: The effectiveness of chemotherapy regimens for breast cancer is highly variable from patient to patient. As a result, a substantial proportion of patients receive toxic drugs from which they fail to benefit, and some patients are delayed in getting the regimen from which they would benefit most.

Topic addressed: Accurate predictors of early response to chemotherapy compounds may prove valuable in determining optimal clinical management of patients with breast cancer. To that purpose, we are developing a method to use positron emission tomography (PET) imaging along with chemotherapy agents labeled with radioactive fluorine (18F), to measure their biodistribution and pharmacokinetics (i.e., how much of each chemotherapy agent goes into tumor and normal organs, how quickly, and for how long). That non-invasively obtained information may then be used to individualize the optimal therapeutic regimen for each patient.

Progress towards specific aims: We have successfully incorporated fluorine into each of two chemotherapy agents used in the treatment of breast cancer, paclitaxel and cyclophosphamide, and have also accomplished their radiosynthesis with the positron-emitting label 18F, creating agents suitable for PET scanning. We have further demonstrated that both [18F]-fluoropaclitaxel and [18F]-cyclophosphamide tracers achieve similar concentrations and distribution times in the blood and organs of mice as do paclitaxel and cyclophosphamide respectively, and that their levels of uptake are not inhibited by treatment doses of unlabeled compound. Furthermore, [18F]-fluoropaclitaxel was found to behave similarly to paclitaxel in response to with cyclosporine, a modulator of chemotherapy resistance mechanisms. We have studied mice with human breast tumors (MCF-7 cell line) grafted in them using microPET (a PET scanner dedicated for small-animal use) imaging to measure the biodistribution and pharmacokinetics of both [18F]-fluoropaclitaxel and [18F]-cyclophosphamide tracers in vivo (inside the body). MicroPET studies have documented the ability to provide quantification of chemotherapy concentrations in tissues through measurements made by this noninvasive imaging method, with these novel [18F]-fluorochemotherapy compounds. We have also found promising results with respect to the predictive value of these measurements for [18F]-fluoropaclitaxel in forecasting response to paclitaxel, of the human breast cancers in the tumor-bearing mice.

Future direction and impact: Our near-future plans involve two main paths of action: 1) to expand the battery of [18F]-labeled radiopharmaceuticals to analogs of other chemotherapy agents commonly used against breast cancer, and 2) to initiate translational studies, gathering pilot data with scans of breast cancer-patients using the new classes of imaging agents developed in our laboratory.


Symposium Abstract (2003)
Background: The effectiveness of breast cancer chemotherapy, traditionally selected on the basis of tumor histology, clinical stage, receptor status, and body surface area is highly variable from patient to patient. A significant obstacle to improving that situation has been that prediction models designed to test chemotherapy agents against tumor cell lines or biopsy specimens in the laboratory have been unsuccessful in identifying many cases of resistance to those agents that occur inside the body. This can result in some patients receiving toxic drugs from which they fail to benefit, and in being delayed in getting the regimen from which they might most benefit. Accurate predictors of early response to chemotherapy compounds may prove valuable in determining optimal clinical management of patients with breast cancer. To that end, we have used positron emission tomography (PET) to visualize chemotherapy agents labeled with radioactive fluorine (18F) to non-invasively assess the distribution pattern and timing of how much of each chemotherapy agent goes into tumors and normal organs, how quickly, and for how long.

Methods: After successfully labeling three chemotherapy agents commonly used to treat breast cancer with 18F, it was determined whether each tracer ([18F]-5-fluorouracil, [18F]-fluorocyclophosphamide, and [18F]-fluoropaclitaxel) had the same distribution patterns in the blood and organs of mice, as the unlabeled agents, and in presence of varying doses of chemotherapy agents. These measurements were made non-invasively with PET imaging and validated by the gold standard method of collecting and analyzing tissues of mice.

Results and Conclusions: The PET tracers had distribution patterns comparable to the unlabeled chemotherapy agents in analyzed tissues, and PET imaging measurements corresponded well with those analyses. The data suggest that PET measures of the tracers can be used to accurately reflect the fate of their corresponding chemotherapy agents. These tracers are thus promising agents for non-invasively predicting the relative amounts of the chemotherapy drug in the tumor and each organ at the dose given for treatment, thereby helping to guide oncologistsí selection of the optimal regimen for each patient.

A non-invasive assay to predict chemotherapy response to 5-fluorouracil using [18F]-5FU and PET
Periodical:Society of Nuclear Medicine
Index Medicus:
Authors: Townsend A, et al., and Silverman DH
Yr: 2004 Vol: Nbr: Abs: Pg:

Predicting chemotherapy response to paclataxel using [18F]-fluoropactataxel and PET.
Periodical:Society of Nuclear Medicine
Index Medicus:
Authors: Hsueh WA, et al., and Silverman DH
Yr: 2004 Vol: Nbr: Abs: Pg: