Chemoprevention research: implications for primary care practice.

The field of cancer prevention research is entering a time of growth and opportunity. This important research is identifying agents that are making a substantial difference by reducing cancer incidence in high-risk populations. Primary care providers are natural partners for this research because of their diversity, commitment to disease prevention, and long-term access to their patient population. Several national chemoprevention trials in breast and prostate cancer are open and seeking to affiliate with primary care providers. Information is provided on this research effort, the development of chemoprevention trials, and how to learn more.

Tamoxifen and fenretinide in women with metastatic breast cancer.

BACKGROUND: Tamoxifen and fenretinide combination therapy has been shown to be an active treatment regimen in metastatic breast cancer patients. This pilot study sought to determine whether the addition of fenretinide to tamoxifen would be associated with antitumor activity in metastatic breast cancer patients who had been previously treated with tamoxifen or who had hormone receptor negative disease. The effect of this therapy on circulating plasma transforming growth factor-beta (TGF-beta) levels and serum lipids was also examined. PATIENTS AND METHODS: Thirty-one patients were treated with tamoxifen (20 mg p.o. daily), and fenretinide (400 mg p.o. daily with a 3-day drug holiday each month). Plasma TGF-beta testing was performed using isoform specific sandwich ELISA. RESULTS: Twenty four of the 31 patients were evaluable for an antitumor response including 14 estrogen receptor (ER) positive patients who had failed prior tamoxifen therapy, seven ER-negative patients, and three hormone therapy naive ER-positive patients. There were no objective antitumor responses; three patients had stable disease for 8, 8, and 24 months. Five patients (16%) discontinued therapy for toxicity (one for grade 3 skin rash and four for abnormal dark adaptation). There was a statistically significant decrease in total cholesterol (median change per patient of -13.5 mg/dl; p = 0.049, a 6.5% decrease), and an increase in HDL levels (median change per patient of +18 mg/dl, p = 0.0001, a 35% increase) with tamoxifen and fenretinide therapy. TGF-beta1 plasma levels were normal in 26 of 28 patients, and no changes in these levels post-treatment were demonstrated. CONCLUSIONS: Tamoxifen and fenretinide therapy is not an active combination in ER negative metastatic breast cancer or in patients whose disease has progressed on tamoxifen. This combination had a beneficial effect on total serum cholesterol and HDL levels with no associated rise in serum triglyceride levels. The 400 mg dose of fenretinide was associated with symptomatic nyctalopia in one-third of patients making it an unsuitable dose for use in breast cancer prevention studies.

Phase I crossover study of paclitaxel with r-verapamil in patients with metastatic breast cancer.

PURPOSE: We conducted a phase I crossover study of escalating doses of both paclitaxel (Taxol; Bristol-Myers, Squibb, Princeton, NJ) and r-verapamil, the less cardiotoxic stereoisomer, in heavily pretreated patients with metastatic breast cancer. PATIENTS AND METHODS: Twenty-nine patients refractory to paclitaxel by 3-hour infusion were treated orally with r-verapamil every 4 hours starting 24 hours before the same-dose 3-hour paclitaxel infusion and continuing for a total of 12 doses. Once the maximum-tolerated dose (MTD) of the combination was determined, seven additional patients who had not been treated with either drug were evaluated to determine whether the addition of r-verapamil altered the pharmacokinetics of paclitaxel. Consenting patients had tumor biopsies for P-glycoprotein (Pgp) expression before receiving paclitaxel and after becoming refractory to paclitaxel therapy. RESULTS: The MTD of the combination was 225 mg/m2 of r-verapamil every 4 hours with paclitaxel 200 mg/m2 by 3-hour infusion. Dose-limiting hypotension and bradycardia were observed in three of five patients treated at 250 mg/m2 r-verapamil. Fourteen patients received 32 cycles of r-verapamil at the MTD as outpatient therapy without developing cardiac toxicity. The median peak and trough serum verapamil concentrations at the MTD were 5.1 micromol/L (range, 1.9 to 6.3), respectively, which are within the range necessary for in vitro modulation of Pgp-mediated multidrug resistance (MDR). Increased serum verapamil concentrations and cardiac toxicity were observed more frequently in patients with elevated hepatic transaminases and bilirubin levels. Hematologic toxicity from combined paclitaxel and r-verapamil was significantly worse compared with the previous cycle of paclitxel without r-verapamil. In the pharmacokinetic analysis, r-verapamil delayed mean paclitaxel clearance and increased mean peak paclitaxel concentrations. CONCLUSION: r-Verapamil at 225 mg/m2 orally every 4 hours can be given safely with paclitaxel 200 mg/m2 by 3-hour infusion as outpatient therapy and is associated with serum levels considered active for Pgp inhibition. The addition of r-verapamil significantly alters the toxicity and pharmacokinetics of paclitaxel.

Phase I/II study of 72-hour infusional paclitaxel and doxorubicin with granulocyte colony-stimulating factor in patients with metastatic breast cancer.

PURPOSE: We conducted a phase I/II trial of concurrently administered 72-hour infusional paclitaxel and doxorubicin in combination with granulocyte colony-stimulating factor (G-CSF) in patients with previously untreated metastatic breast cancer and bidimensionally measurable disease. PATIENTS AND METHODS: We defined the maximum-tolerated dose (MTD) of concurrent paclitaxel and doxorubicin administration and then studied potential pharmacokinetic interactions between the two drugs. Forty-two patients who had not received prior chemotherapy for metastatic breast cancer received 296 total cycles of paclitaxel and doxorubicin with G-CSF. RESULTS: The MTD was determined to be paclitaxel 180 mg/m2 and doxorubicin 60 mg/m2 each by 72-hour infusion with G-CSF. Diarrhea was the dose-limiting toxicity (DLT) of this combination, with three of three patients developing abdominal computed tomographic (CT) scan evidence of typhlitis (cecal thickening) at the dose level above the MTD. All patients developed grade 4 neutropenia (absolute neutrophil count [ANC] < 500 microL), generally less than 5 days in duration. This combination was generally safely administered at dose levels at or below the MTD. The overall response rate was 72% (28 of 39 patients; 95% confidence interval [CI], 55% to 85%), with 8% complete responses (CRs) (three of 39; 95% CI, 2% to 21%) and a median response duration of 9 months. The median overall survival time for all patients is 23 months, with a median follow-up duration of 28 months. Pharmacokinetic studies showed that administration of paclitaxel and doxorubicin together by 72-hour infusion did not affect the steady-state concentrations of either drug. CONCLUSION: Concurrent 72-hour infusional paclitaxel and doxorubicin can be administered safely, but is associated with significant toxicity. The overall response rate of this combination in untreated metastatic breast cancer patients is similar to that achieved with other doxorubicin-based combination regimens. The modest complete response rate achieved suggests that this schedule of paclitaxel and doxorubicin administration does not produce significant additive or synergistic cytotoxicity against breast cancer.

Phase I study of paclitaxel in combination with cyclophosphamide and granulocyte colony-stimulating factor in metastatic breast cancer patients.

PURPOSE: In vitro data suggest that prolonged exposure to paclitaxel enhances breast cancer cytotoxicity. Our objective in this phase I study was to determine the tolerability of paclitaxel administered by 72-hour continuous intravenous (i.v.) infusion (CIVI) in combination with high-dose cyclophosphamide and granulocyte colony-stimulating factor (G-CSF) in the ambulatory setting to metastatic breast cancer patients. PATIENTS AND METHODS: Paclitaxel was administered over 72 hours by CIVI and cyclophosphamide was given daily by i.v. bolus on days 1, 2, and 3, followed by G-CSF every 21 days. The availability of ambulatory infusion pumps and paclitaxel-compatible tubing permitted outpatient administration. RESULTS: Fifty-five patients with metastatic breast cancer who had been previously treated with a median of two prior chemotherapy regimens were entered onto the study. Dose-limiting toxicity of grade 4 neutropenia for longer than 5 days and grade 4 thrombocytopenia occurred in three of five patients treated with paclitaxel 160 mg/m2 CIVI and cyclophosphamide 3,300 mg/m2 followed by G-CSF. The maximum-tolerated dose (MTD) was paclitaxel 160 mg/m2 CIVI and cyclophosphamide 2,700 mg/m2 in divided doses with G-CSF. Nonhematologic toxicities were moderate and included diarrhea, mucositis, and arthalgias. Although hemorrhagic cystitis developed in six patients, recurrence was prevented with i.v. and oral mesna, which permitted continued outpatient delivery. One hundred seventy-four cycles were safely administered in the ambulatory setting using infusional pumps and tubing. Objective responses occurred in 23 (one complete and 22 partial) of 42 patients with bidimensionally measurable disease (55%; 95% confidence interval, 38% to 70%), with a response rate of 73% (11 of 15) seen at the highest dose levels. CONCLUSION: Paclitaxel by 72-hour CIVI with daily cyclophosphamide followed by G-CSF can be administered safely in the ambulatory setting, has acceptable toxicity, and is an active regimen in the treatment of metastatic breast cancer.

A phase I study of sequential versus concurrent interleukin-3 and granulocyte-macrophage colony-stimulating factor in advanced breast cancer patients treated with FLAC (5-fluorouracil, leucovorin, doxorubicin, cyclophosphamide) chemotherapy.

Cumulative thrombocytopenia is a dose-limiting toxicity of dose-intensive chemotherapy for advanced breast cancer. In this phase I study, we have studied the hematologic toxicity associated with sequential interleukin-3 (IL-3) and granulocyte-macrophage colony-stimulating factor (GM-CSF; molgramostim) administration after multiple cycles of FLAC (5-fluorouracil, leucovorin, doxorubicin, cyclophosphamide) chemotherapy compared with that after concurrent cytokine administration or to each cytokine administered alone. Ninety-three patients with advanced breast cancer were treated with five cycles of FLAC chemotherapy and either IL-3 alone, GM-CSF alone, sequential IL-3 and GM-CSF administered by schedule A (5 days of IL-3 followed by 10 days of GM-CSF) or schedule B (9 days of IL-3 followed by 6 days of GM-CSF), or concurrent administration of IL-3 and GM-CSF for 15 days. Cohorts of patients were treated with one of four dose levels of IL-3 (1,2.5, 5, and 10 micrograms/kg) administered subcutaneously for each schedule of cytokine administration. The GM-CSF dose in all schedules was 5 micrograms/kg/day. Sequential IL-3 and GM-CSF (schedule B) was associated with higher platelet nadirs, shorter durations of platelet counts less than 50,000/microL, and the need for fewer platelet transfusions over five cycles of FLAC chemotherapy compared with concurrent cytokines, sequential IL-3 and GM-CSF schedule A, and GM-CSF alone. Concurrent IL-3 and GM-CSF was associated with unexpected platelet toxicity. The duration of granulocytopenia after FLAC chemotherapy was significantly worse with IL-3 alone compared with each of the GM-CSF-containing cytokine regimens. Although no cycle 1 maximum tolerated dose for IL-3 was defined in this study, 5 micrograms/kg was well tolerated over multiple cycles of therapy and is recommended for future studies. The data from this phase I study suggest that sequential IL-3 and GM-CSF with IL-3 administered for 9 days before beginning GM-CSF may be superior to shorter durations of IL-3 administered sequentially with GM-CSF, to concurrent IL-3 and GM-CSF, and to either colony-stimulating factor alone in ameliorating the cumulative hematologic toxicity associated with multiple cycles of FLAC chemotherapy. Additional studies of sequential IL-3 and GM-CSF are warranted.

Effect of R-verapamil on the pharmacokinetics of paclitaxel in women with breast cancer.

PURPOSE: To study the effect of the multidrug-resistance reversal agent R-verapamil on the pharmacokinetic behavior of paclitaxel. METHODS: Six women with breast cancer who received paclitaxel as a 3-hour infusion with and without R-verapamil were monitored with frequent plasma sampling up to 24 hours postinfusion. Paclitaxel concentrations were measured using a reverse-phase high-pressure liquid chromatography assay. RESULTS: Concomitant administration of R-verapamil resulted in a decrease in mean (+/- SD) paclitaxel clearance from 179 +/- 67 mL/min/m2 to 90 +/- 34 mL/min/m2 (P < .03) and in a twofold increase in paclitaxel exposure (area under the curve [AUC]). The mean end-infusion paclitaxel concentration was also twofold higher: 5.1 +/- 1.8 mumol/L versus 11.3 +/- 4.1 mumol/L (P < .03). CONCLUSION: The alteration in paclitaxel pharmacokinetics when paclitaxel and R-verapamil are coadministered complicates the interpretation of response and toxicity data from clinical trials of this drug combination.

Chemotherapy-induced changes in uterine volume: CT study.

OBJECTIVE: The goal of this study was to quantify changes in uterine volume during and after chemotherapy for breast cancer. MATERIALS AND METHODS: Fifty-eight patients (mean age 42 years) with advanced breast cancer undergoing treatment with FLAC (5-fluorouracil, leucovorin, Adriamycin, and cyclophosphamide) were studied with serial pelvic CT. The transverse, anteroposterior, and sagittal measurements of the uterus were taken in each of the CT scans. The uterine volumes were calculated, normalized to baseline volumes, and graphically displayed for each patient. The temporal changes in uterine volume were correlated to the dates of chemotherapy administration and menstrual status. RESULTS: There was a striking and consistent loss of uterine volume with chemotherapy. This phenomenon was observed in 55 (95%) of the 58 patients. The mean minimum uterine volume was 58.3 +/- 20.2% compared with the baseline. This loss of uterine volume began after the administration of the first cycle of chemotherapy and progressed during subsequent cycles. It coincided with loss of normal menses in all patients. After completion of chemotherapy, there was a recovery of the uterine volume in 12 of the 16 patients who had follow-up CT. The mean recovery uterine volume was 108.4 +/- 49.8% of baseline. Menses recovered in four of these. CONCLUSION: Chemotherapy causes loss of uterine volume, which usually recovers after the withdrawal of chemotherapy in premenopausal women. This change should not be mistaken for a therapeutic response of primary or secondary malignancy within the uterus or be confused with subsequent uterine changes due to tamoxifen.

A pilot study of amiodarone with infusional doxorubicin or vinblastine in refractory breast cancer.

Increasing evidence suggests that P-glycoprotein (Pgp) expression can mediate drug resistance in refractory breast cancer. We studied 33 patients with refractory breast cancer enrolled in a pilot study of oral amiodarone as a Pgp antagonist given in combination with infusional doxorubicin or vinblastine. Whenever possible, tumors were biopsied and Pgp expression was assayed. Patients received either 60 mg/m2 doxorubicin over 96 h or 8.5 mg/m2 vinblastine over 120 h by continuous intravenous infusion. Beginning with the second cycle of chemotherapy, 600-800 mg amiodarone was given orally each day. Patients who experienced toxicity due to amiodarone but were responding to chemotherapy were placed on quinidine. Partial responses were observed in 9 of 33 patients on study and were sometimes observed after the first cycle of chemotherapy, before amiodarone was given, suggesting that some patients may have responded to treatment because of the infusional schedule. Toxicities were primarily the known side effects of the antineoplastic agents and of amiodarone. The major amiodarone toxicity was gastrointestinal, with nausea, vomiting, anorexia, or diarrhea being noted in 21 patients. Biopsy samples were obtained from 29 patients and in 21 cases, viable tumor tissue was present and the results were interpretable. Of the 21 samples, 9 had Pgp expression as determined by immunohistochemical staining; 12 were considered negative. The presence of Pgp expression was associated with an acceleration of the time to treatment failure. Whereas normal-tissue toxicities related to the combination of a Pgp antagonist with chemotherapy were not observed, amiodarone was associated with too many untoward effects to be utilized as a drug resistance-reversing agent.

Transforming growth factor-beta1 circulates in normal human plasma and is unchanged in advanced metastatic breast cancer.

A method has been developed to determine true plasma transforming growth factor beta (TGF-beta) levels by using the platelet alpha granule-specific marker, platelet factor 4, to correct for the TGF-beta contributed by platelets degranulated ex vivo. TGF-beta levels were measured on acid-ethanol extracts of human plasma using isoform-specific sandwich enzyme-linked immunosorbent assays. Normal human subjects had 4.1 +/- 2.0 ng/ml TGF-beta1 (range, 2.0-12.0; n = 42), <0.2 ng/ml TGF-beta2, and <0.1 ng/ml TGF-beta3 in their plasma. There were no significant changes with age or with hormonal status, but any given individual showed fluctuations of up to 3-fold in measured plasma TGF-beta levels due to unknown factors. Of 28 patients with advanced metastatic breast cancer, 2 had greatly elevated TGF-beta1 levels, while the rest were in the normal range. The presence of physiologically significant levels of TGF-beta1 in the plasmas of normal human subjects may indicate previously unsuspected endocrine roles for this peptide, while TGF-beta2 and TGF-beta3 appear to act only in a local autocrine/paracrine fashion.

A dose intensity study of FLAC (5-fluorouracil, leucovorin, doxorubicin, cyclophosphamide) chemotherapy and Escherichia coli-derived granulocyte-macrophage colony-stimulating factor (GM-CSF) in advanced breast cancer patients.

BACKGROUND: It has been demonstrated that granulocyte-macrophage colony-stimulating factor (GM-CSF) can ameliorate chemotherapy-induced neutropenia. The extent to which GM-CSF can increase the actual delivered dose intensity of combination chemotherapy over multiple cycles of therapy to patients with advanced breast cancer has not been well defined. We conducted a phase I/II study of dose-intensive FLAC chemotherapy (5-fluorouracil, leucovorin, doxorubicin, cyclophosphamide) in combination with GM-CSF in patients with locally advanced and metastatic breast cancer to study the acute and cumulative toxicities, anti-tumor activity and dose-intensity achievable with this regimen. METHODS: Eighty-one patients with newly diagnosed stages IIB, III and IV breast cancer who were previously untreated with chemotherapy and who had measurable disease received multiple cycles of FLAC chemotherapy plus E. coli-derived GM-CSF administered every three weeks. RESULTS: FLAC plus GM-CSF as associated with significant cumulative hematologic toxicity. Ninety-eight percent of patients developed grade 4 neutropenia; 29% of all cycles administered required hospitalization for fever and neutropenia; 41% and 22% of cycles required red blood cell and platelet transfusions, respectively. Other significant toxicities with E. coli-derived GM-CSF included mild to moderate first dose effects (hypotension, dyspnea, abdominal cramping) in 30% of patients; late occurring anaphylactoid reactions in 11% of patients; and vascular thromboses. The average delivered dose intensities over all cycles were cyclophosphamide, 210 mg/m2/week; doxorubicin, 14.8 mg/m2/week and 5-fluorouracil, 342 mg/m2/week. The overall clinical response rates were 100% and 83% for LABC and metastatic patients, respectively. There were 23% (6/26) pathologic CR's in the LABC patients given neoadjuvant FLAC and 22% (12/54) clinical CR's in the stage IV patients. The median survival of the LABC patients has not been reached (> 26 months) and is 30 months for the stage IV patients. CONCLUSIONS: The administration of multiple cycles of FLAC plus E. coli-derived GM-CSF therapy is associated with cumulative, dose-limiting myelosuppression, especially thrombocytopenia, as well as significant clinical toxicity. A modest increase in FLAC dose intensity over the starting doses was achievable with the addition of GM-CSF. FLAC chemotherapy has substantial antitumor activity in the treatment of advanced breast cancer. The potential usefulness of FLAC plus GM-CSF must be balanced by its considerable cost and alteration in patients' quality of life due to toxicity. Combination hematopoietic growth factor strategies may be able to reduce the toxicity of FLAC and to allow further increase dose intensity.

Phase I study of high-dose piroxantrone with granulocyte colony-stimulating factor.

PURPOSE: We performed a phase I trial of piroxantrone with and without granulocyte colony-stimulating factor (G-CSF) to determine whether the use of this cytokine would enable us to increase the dose-intensity of piroxantrone. PATIENTS AND METHODS: Thirty-eight patients received 121 courses of piroxantrone administered once every 21 days. Initial patient cohorts received piroxantrone alone starting at 150 mg/m2 and the dose was escalated in subsequent patients until dose-limiting toxicity (DLT) was reached. Patient cohorts then received escalating doses of piroxantrone starting at 185 mg/m2 administered with G-CSF beginning day 2. RESULTS: Dose-limiting neutropenia occurred in three of six patients treated with 185 mg/m2 piroxantrone; the maximum-tolerated dose (MTD) of piroxantrone alone was 150 mg/m2. Three of six patients treated with piroxantrone and G-CSF exhibited dose-limiting thrombocytopenia at 445 mg/m2; the MTD of piroxantrone with G-CSF was thus 355 mg/m2. Seven patients developed symptomatic congestive heart failure (CHF) at cumulative piroxantrone doses ranging from 855 to 2,475 mg/m2 and two have died of cardiotoxicity. Of these patients, six of seven had previously received doxorubicin. Other nonhematologic toxicity was mild. CONCLUSION: The use of G-CSF results in a more than twofold increase in the MTD of piroxantrone. However, symptomatic cardiotoxicity is prominent, especially in patients who have received prior treatment with anthracyclines.

Pharmacokinetics of piroxantrone in a phase I trial of piroxantrone and granulocyte-colony stimulating factor.

Piroxantrone is an anthrapyrazole derivative with broad antitumor activity in vitro. In previous phase I trials, the dose-limiting toxicity of this agent was myelosuppression. Therefore, a phase I and pharmacokinetic study of a 1-h infusion of piroxantrone in combination with granulocyte-colony stimulating factor was conducted. In this article, we report the results of the pharmacokinetic analysis. Thirty-seven patients were studied over a dosage range of 150 to 555 mg/m2. The plasma elimination of piroxantrone was biexponential with a mean (+/- SD) t1/2 alpha of 3.2 +/- 2.7 min and a mean (+/- SD) t1/2 beta of 82 +/- 92 min. Clearance was 840 +/- 230 ml/min/m2. A limited sampling strategy was developed to allow the estimation of total drug exposure (area under the plasma concentration-time curve) from the plasma piroxantrone concentrations at 30, 60, and 120 min after the start of the infusion. The pharmacokinetic behavior of a presumed piroxantrone metabolite not previously described in plasma was also characterized. Based on in vitro cytotoxicity studies with partially purified extract of this compound, we do not believe that it contributes to the antitumor effects of piroxantrone at the concentrations observed in plasma. Finally, piroxantrone elimination was linear over the nearly 4-fold dose range studied, indicating that when dose adjustments are made, systemic drug exposure will remain predictable.

Oncology nurse and physician attitudes toward aggressive cancer treatment.

Oncology nurses (n = 57) and oncology physicians (n = 40) employed in the same hospital completed the Cancer Attitude Questionnaire. Significant differences (p < or = .001) between nurse and physician attitudes were evident on the two major subscales addressing the issues of (a) aggressiveness of treatment/de-emphasis of socioemotional aspects of care and (b) the importance of patient-family attitudes. Physicians' attitudes were significantly more favorable on the first subscale; nurses' attitudes were significantly more favorable on the second. Implications of these findings as indicators of potential nurse-physician conflict are discussed.

Phase II study of fazarabine (NSC 281272) in patients with metastatic colon cancer.

Fazarabine (Arabinofuranosyl-5-azacytosine) is a synthetic pyrimidine nucleoside which combines the arabinose sugar of cytosine arabinoside with the triazine base of 5-azacytidine. It has demonstrated activity against a variety of human solid tumor xenografts including colon, lung and breast cancers. Eighteen patients with refractory metastatic colon cancer were enrolled in a phase II trial of fazarabine. The drug was administered as a 72 hr continuous infusion every 3-4 weeks; the starting dose was 2 mg/m2/hr as established in a previous phase I study. The major toxicity was neutropenia, as predicted from the phase I study. The median time to nadir for cycle 1 was 20 days, with a median granulocyte count of 437/microliters (range 36-1600/microliters); recovery was within 2-4 days, with only one incidence of fever and neutropenia in 42 cycles. Especially noted for their absence were thrombocytopenia, nausea, vomiting and stomatitis. No objective clinical responses were seen; one patient had stabilization of rapidly growing liver metastases for a period of 7 months. In view of fazarabine's narrow range of toxicities, future dose intensification trials utilizing fazarabine in combination with hematopoietic growth factors are worthy of consideration.

Nursing management of the patient receiving Taxol therapy.

Taxol is an important new antitumor agent with demonstrated efficacy in ovarian and breast cancer. Toxicities identified, including cardiac, hypersensitivity reactions, and neurologic, require careful nursing assessment for management, Additional toxicities may be identified as Taxol is combined with other chemotherapeutic agents. Studies to determine the most effective dose and schedule are ongoing. The current evaluation of this new drug presents an important opportunity for nurses to contribute to its development through both clinical and research endeavors. Such contributions will facilitate the optimal nursing care of patients treated with Taxol.

Therapy of patients with metastatic breast cancer with 5-fluorouracil, leucovorin and carboplatin.

Thirty-four women with metastatic breast cancer were treated at the National Cancer Institute of the National Institutes of Health, with a regimen of leucovorin (L), 500 mg/m2 i.v. over 30 min, followed in 1 h by 5-fluorouracil (5-FU), 375 mg/m2 i.v. bolus on days 1-5, and carboplatin (CBDCA), 50-100 mg/m2 i.v. bolus on days 2-4, every 28 days. All patients had received previous combination chemotherapy with at least one regimen (29 patients with 5-FU-containing regimens). CBDCA, 100 mg/m2 on days 2-4, resulted in grade 4 neutropenia in 10 out of 11 patients associated with sepsis in all 10 patients. CBDCA, 75 mg/m2 (seven patients) and 50 mg/m2 (15 patients), resulted in grade 4 neutropenia in six and eight patients, and neutropenic sepsis in five and two cases, respectively. Grade 4 thrombocytopenia occurred in 10, five and two patients receiving 100, 75 and 50 mg/m2 of CBDCA, respectively. Other toxicities included grade 3/4 mucositis in 18 patients and grade 3/4 diarrhea in 10 patients. Twenty nine patients were evaluable for response, with one pathologic complete response (3%), two partial responses (6%), 18 stable disease (53%) and eight (24%) progressive disease. Sites of response included bone, viscera and soft tissue. The median time from entry on study to progression, for responders, was 15 months. When platinum-DNA adduct formation in peripheral white blood cells was analyzed in 27 patients at 24 h after drug administration, a significant correlation between adduct level and CBDCA cumulative dose was found.(ABSTRACT TRUNCATED AT 250 WORDS)