Reflections on the development of resistance during therapy for advanced breast cancer. Implications of high levels of activity of docetaxel in anthracycline-resistant breast cancer patients

Anthracyclines play a central role in the treatment of breast cancer. They are perhaps the most active single agents available for the treatment of this disease. For patients with primary breast cancer for whom chemotherapy would be appropriate, anthracyclines are often incorporated in adjuvant regimens. For patients with endocrine-therapy-resistant metastatic disease, anthracyclines are nearly always included in the first or second-line therapeutic regimes. Unfortunately, despite the fact that anthracyclines and anthracycline-containing regimens can achieve impressive objective response rates, this impressive activity has not translated into a great improvement in disease-free survival for patients with early stage breast cancer or overall survival for patients with metastatic disease. The addition of anthracyclines to adjuvant therapy regimens in general only modestly improves their efficacy and the use of anthracyclines in metastatic disease does not cure these patients. The clinical utility of anthracyclines would be greatly improved if we could predict which patients would be anthracycline resistant, interfere with the development or expression of anthracycline resistance and predict which anticancer agents would be non-cross-resistant with anthracyclines. Some progress is being made in all these areas. Preclinical studies have identified several intracellular processes that are perturbed by anthracyclines, or that may modulate anthracycline sensitivity of cells. These processes include topoisomerase II activity, drug and toxin transmembrane pumps, intracellular detoxification systems (such as that related to gluatathione), stress-related proteins and apoptotic mechanisms. Although measurement of the components of these systems has not yet shown clinical utility in breast cancer, some preclinical work and exploratory studies with small numbers of patients suggest that we may in the future be able to predict which patients will respond or be resistant to anthracyclines. An important avenue of work is the identification of anticancer agents that are non-cross-resistant with anthracyclines in breast cancer patients. These agents would be particularly valuable in patients with metastatic disease who had progressed while on anthracyclines. In general, such patients have a very poor prognosis with a median survival of less than 1 year. Also of importance in non-cross-resistant agents is that they might be used in combination with anthracyclines in regimens with very high response rates. Recently completed work suggests that taxoids might be such agents. This finding opens up exciting possibilities in the treatment of metastatic disease and, even more importantly, in the adjuvant arena. The identification of agents that clinically are non-cross-resistant with anthracyclines depends on the careful interpretation of clinical trial data. Unfortunately, a number of definitions of anthracycline resistance have been used in the medical literature. These range from very weak definitions, which include many patients with only prior anthracycline exposure (for example, in an adjuvant regimen), to more biologically and clinically appropriate definitions, such as documented progression while receiving an anthracycline. This distinction is very important because it is clear that many patients who have relapsed after an anthracycline-based adjuvant therapy will respond to anthracyclines for metastatic disease and are, therefore, not truly anthracycline resistant. In this regard, the recently completed trials with docetaxel in patients with rigorously defined anthracycline resistance are particularly provocative. These trials show that docetaxel maintains much of its excellent first-line levels of efficacy in patients with anthracycline-resistant breast cancer. Agents with this type of maintenance of efficacy in anthracycline-resistant tumours may find immediate utility in this clinical scenario as single agents but, more importantly, have gr

Docetaxel (Taxotere) for the treatment of anthracycline-resistant breast cancer

Until the introduction of the taxoids, docetaxel (Taxotere; Rhône-Poulenc Rorer, Antony, France) and paclitaxel (Taxol; Bristol-Myers Squibb Oncology, Princeton, NJ), in the 1990s, anthracyclines were widely recognized as the best single agents for the treatment of breast cancer. However, even when anthracyclines are used in combination regimens with response rates of over 50%, including complete responses in 17% of patients, few women (3%) with metastatic disease remain disease free at 5 years after treatment. The low level of sustained responses is largely due to the phenomenon of drug resistance. Anthracycline resistance often involves multidrug resistance efflux mechanisms, but also can involve factors affecting topoisomerase II and apoptosis. When combining other cytotoxic agents with anthracyclines, it is of value to use non-cross-resistant drugs so that the induction of anthracycline-resistance mechanisms does not also affect the efficacy of other agents in the combination therapy. Clinical studies have shown that docetaxel, which is highly active against metastatic breast cancer as a single agent, has a high level of non-cross-resistance with anthracyclines. The overall response rate to docetaxel monotherapy in patients with anthracycline-resistant or refractory metastatic disease has been shown to be 41%. The response rate to first-line docetaxel monotherapy for metastatic breast cancer has been shown to be 61%, suggesting that two thirds of the activity of docetaxel is retained in anthracycline-resistant disease. Treatment with a simultaneous combination of docetaxel and doxorubicin has been found to be very active, with a response rate of 89%, and trials to exploit the lack of cross-resistance between these agents, in sequential regimens and adjuvant therapies, are under way.

Lack of prognostic value of cathepsin D levels for predicting short term outcomes of breast cancer patients

The value of cathepsin D determinations done on tumor cytosols in evaluating the prognosis of breast cancer patients has been debated in the literature. Our previous work suggested that cathepsin D determinations were not of prognostic value, but in that study we used immunoblotting and immunohistochemical methods rather than the more widely used double antibody immunoradiometric (IRMA) assay for measuring cathepsin levels. Here we report our results determining cathepsin D using components of a commercially available IRMA system on a large patient sample (n = 1984). Reagents from a commercially available IRMA kit were used to analyze cathepsin D levels in the cytosols of 1984 patients with breast cancer. All patients had invasive breast cancer with known tumor size and with some axillary nodes pathologically examined. Only patients with T1 and T2 tumor sizes were included. Median follow-up was 37 months. The hypothesis that high cathepsin D levels correlated with poorer outcome (poorer DFS or OS) was not confirmed, either in all patients, or in node-positive or node-negative subsets. Only in patients treated with adjuvant therapy were higher cathepsin D levels correlated with negative outcome (worsened OS, but not DFS), although given the large number of subsets analyzed this correlation may be spurious. Multivariate analyses using interaction terms did not support the concept that high cathepsin D levels correlate with resistance to adjuvant therapy. In this study evaluating the value of cathepsin D using components from a kit widely used for measuring cathepsin D levels, we conclude that cathepsin D is of doubtful value in predicting risk of early relapse or death for patients with newly diagnosed invasive breast cancer.

The international experience with docetaxel in the treatment of breast cancer

The extensively studied agent docetaxel (Taxotere) has shown marked clinical activity in the treatment of anthracycline-resistant breast cancer. Phase I trials indicate that toxicities, such as mucositis and neutropenia, limit the administration of docetaxel to shorter perfusion schedules. Pharmacokinetic studies have shown that docetaxel's clearance by hepatic metabolism is correlated with a marked increase in risk of toxicity in patients with impaired liver function. Nevertheless, studies of docetaxel as front-line therapy for breast cancer were initiated because of its good activity against tumors in early studies and its close relationship to paclitaxel (Taxol), an agent with proven efficacy. Phase II studies have demonstrated excellent activity for docetaxel as a single agent, with an overall response rate of 61% in trials of a 100-mg/m2 dose. A phase III study is currently comparing docetaxel with paclitaxel as single-agent therapy. Docetaxel is expected to provide a better response rate but a higher incidence of neutropenia. The agent shows promise in adjuvant therapy, with very high response rates in anthracycline-resistant patients. Preliminary results of tests using docetaxel in combination with doxorubicin show high objective response rates but low complete response rates; early results suggest that this combination may have some advantages over paclitaxel/doxorubicin.

Treatment of patients resistant to anthracycline therapy

The results of three multicentre phase II trials in which docetaxel (Taxotere) was used in previously treated patients with metastatic breast cancer resistant to anthracyclines or anthracenediones are summarized here. Docetaxel was given to a total of 134 patients who had evidence of disease progression while receiving anthracyclines or anthracenediones for metastatic disease or had relapsed during adjuvant therapy which included these agents. The overall response rate (ORR) across the three studies was 41% in an intent-to-treat analysis. The median duration of response varied from 24 to 28 weeks between studies and the median survival varied from 9 to 12 months. The response rate was well maintained in evaluable patients with visceral metastases (ORR 43%), or multiple (> 2) sites of disease (ORR 48%). These response rates are the highest ever reported for a single agent in patients with anthracycline-resistant disease. The recommended dose and schedule for docetaxel (100 mg/m2 intravenously over 1-h every 3 weeks), which was used in all three studies, was found to be well tolerated, with neutropenia as the most common toxicity (grade 4 in 90% of patients) and febrile neutropenia requiring hospitalisation occurring in only 4% of cycles of therapy.

A computer program to assist in making breast cancer adjuvant therapy decisions

This report describes a computer program designed to assist health care professionals in making projections of the average benefit of systemic adjuvant therapy for individual breast cancer patients. It requires as input patient age (used to make projections of natural mortality), an estimate of breast cancer-related mortality at 5 years (used to make projections of breast cancer-specific mortality), and the proportional risk reduction for breast cancer mortality expected for the adjuvant therapy (with included tables from the Early Breast Cancer Trialist's 1992 meta-analysis). The program uses life table analytical techniques to make projections of outcome in three scenarios: that the breast cancer never occurred, that the breast cancer patient received definitive local therapy but no adjuvant systemic therapy, and that the patient received adjuvant therapy. The outcome projections are given for total, natural (non-breast cancer-related), and breast cancer-related mortality at several time points and also of total remaining life expectancy. These estimates are currently widely made by clinicians by nonnumerical techniques. Computer-based tools can serve as valuable aids in physician and patient education and in the process of informed decision making.

A computer based program to assist in adjuvant therapy decisions for individual breast cancer patients

This paper describes a personal computer based tool to aid in decision making about whether a woman should receive adjuvant therapy for breast cancer. This tool can assist in engaging women with primary breast cancer in the discussion about: 1) her risk of breast cancer related mortality if she receives only local control measures, but no systemic adjuvant therapy, 2) how much receiving adjuvant therapy may reduce this risk, and 3) what the impact of receiving the adjuvant systemic therapy is in terms of survival. The tool utilizes life table analytical techniques to project outcomes after entry of patient age (used to calculate natural mortality rates), estimated risk of breast cancer related mortality (with a help tool allowing the physician to use estimates based on national database information), and estimate of the efficacy of adjuvant chemotherapy (with included tables of estimates based on the Early Breast Cancer Trialists' meta-analysis). Computer based tools can serve as valuable aids in patient and physician education, and the process of informed decision making.

Phase II trial of docetaxel in advanced anthracycline-resistant or anthracenedione-resistant breast cancer

PURPOSE

The purpose of this study was to evaluate the clinical efficacy and safety of docetaxel in patients with metastatic breast cancer (MBC) resistant to doxorubicin or mitoxantrone.

PATIENTS AND METHODS

Docetaxel 100 mg/m2 was administered as a 1-hour intravenous (IV) infusion every 3 weeks to 42 patients registered at four centers. Patients must have received at least one but no more than two prior chemotherapy regimens for MBC (in addition to any prior adjuvant therapy). One of the regimens for metastatic breast cancer must have included an anthracycline or anthracenedione and the cancer must have progressed on that regimen.

RESULTS

Objective responses were seen in 20 of 35 assessable patients (three complete responses [CRs] and 17 partial responses [PRs]), for an objective response rate of 57% (95% confidence interval [CI], 39% to 74%) and in 21 of 42 registered patients (50% response rate [RR]; 95% CI, 34% to 66%) entered onto the trial. The median response duration was 28 weeks. The most common toxicity in this study was grade 4 neutropenia, which occurred in 95% of patients. Other clinically significant nonhematologic side effects included stomatitis, skin reactions, neurosensory changes, asthenia, and fluid retention. Patients who received dexamethasone premedication had a later onset of fluid retention than those who did not receive dexamethasone (onset at a median cumulative docetaxel dose of 503 mg/m2 and 291 mg/m2, respectively).

CONCLUSION

Docetaxel at this dose and schedule has a high level of antitumor activity in patients with treatment-refractory advanced breast cancer, and appears to be one of the most active agents for the treatment of this patient population.

Taxoids: effective agents in anthracycline-resistant breast cancer

The results of recent clinical trials have shown that docetaxel (Taxotere; Rhône-Poulenc Rorer, Antony, France), like paclitaxel (Taxol; Bristol-Myers Squibb Oncology, Princeton, NJ), has high levels of activity in patients with anthracycline-resistant breast cancer. Agents that are at least partially non-cross-resistant with anthracyclines are especially promising for the treatment of breast cancer; the taxoids (docetaxel and paclitaxel) are such agents. Although preclinical evaluations shows clear instances of strong cross-resistance (particularly in cells lines expressing the P-glycoprotein, multidrug resistance), high response rates have been reported in patients with prior anthracycline exposure and/or anthracycline resistance. Phase I studies of anthracycline and taxoid combinations have been conducted. Excellent response rates have been noted in some of these studies. In some studies using regimens combining doxorubicin and paclitaxel, unanticipated toxicities have occurred, such as typhlitis, as well as congestive heart failure at lower than expected cumulative doses of doxorubicin. Phase II and III studies of regimens including both anthracyclines and taxoids have been initiated. Docetaxel and paclitaxel appear to be valuable agents for use in anthracycline-resistant breast cancer patients, and may find a place in anthracycline-containing combination regimens.

Anthracycline resistance in breast cancer: clinical applications of current knowledge

Anthracyclines are highly effective antineoplastic agents for the treatment of breast cancer. Nevertheless, essentially all breast cancer patients have tumours which are intrinsically resistant or which develop resistance during the course of therapy. Clinical trials provide indirect information on the nature of anthracycline resistance and work in the basic sciences has demonstrated molecular mechanisms which play a role. Initial clinical attempts to exploit and translate these mechanisms to predict, and interfere with, anthracycline resistance have met with mixed success, and have not yet led to accepted clinical applications.

The c-erbB-2 proto-oncogene as a prognostic and predictive marker in breast cancer: a paradigm for the development of other macromolecular markers--a review

Seven years after the initial studies of the prognostic value of the proto-oncogene c-erbB-2 in breast cancer, its role is still being defined. The interpretation of studies on the use of this gene and its protein product in prognostic and predictive tests for breast cancer is complicated by multiple methodologies and the inherent difficulties in the studies. The work has moved beyond the stage at which small studies with short follow-up (useful for hypothesis generation) are of value, to the stage in which large studies with sufficient statistical power to find significant correlations are central. These larger studies do not lend support for the use of c-erbB-2 in the evaluation of axillary-node-negative patients, the group of breast cancer patients for whom refinement of prognostic estimates is now most important. There are, however, hints that c-erbB-2 may have value in predicting response to certain treatments, though the studies so far are too few, often too small and too conflicting to reliably confirm this.

Review of docetaxel (Taxotere), a highly active new agent for the treatment of metastatic breast cancer

Phase II studies have demonstrated that docetaxel (Taxotere; Rhône-Poulenc Rorer, Collegeville, PA) is one of the most active single agents in the treatment of metastatic breast cancer. The overall response rate as front-line therapy for metastatic disease was 59% (95% confidence interval, 51% to 67%) in five phase II trials (four of which were multicenter) when 100 mg/m2 docetaxel was infused over 1 hour every 3 weeks. In the three phase II trials reported to date of patients with metastatic cancer who had failed previous frontline therapy, 100 mg/m2 docetaxel infused over 1 hour every 3 weeks produced an objective response rate of 49% (95% confidence interval, 40% to 58%). Two of these trials specifically included patients who had progressed while receiving either an anthracycline or an anthracenedione; the overall response rate in this subset of 83 patients was 48%. The most significant acute toxicity noted in these trials was neutropenia. Grade 4 neutropenia occurred in the majority of patients but rarely resulted in treatment delays. Hypersensitivity reactions also were common in nonpremedicated patients, but were rare after the institution of premedication with antihistamines and/or glucocorticoids. A novel toxicity observed in many patients was fluid retention syndrome, with onset at a median of four to five cycles. The fluid retention was of noncardiac or renal origin, was slowly progressive with additional cycles of therapy, was reversible after cessation of the drug, and could be largely ameliorated by oral diuretics and glucocorticoid premedication. Phase III studies to further define docetaxel's role in the treatment of breast cancer are now under way.

Prediction of axillary lymph node status in breast cancer patients by use of prognostic indicators

BACKGROUND

If axillary lymph node status of breast cancer patients could be accurately predicted from basic clinical information and from characteristics of their primary tumors, then many patients could be spared axillary lymph node dissection. Tumor size alone does not allow the identification of groups with very low or high risk of being axillary node positive.

PURPOSE

Our goal was to investigate the possibility of using prognostic indicators to predict axillary node status of patients with primary breast cancer.

METHODS

Data from 26,683 patients from the National Breast Cancer Tissue Resource were used in this study. Patients in this dataset were randomly assigned to a training set (patient information used to construct predictive models) or a validation set (patient information used to prospectively evaluate predictive models). The records of a total of 11,964 case patients that had complete prognostic factors and pathologic data were analyzed: 5963 patients in the training set and 6001 patients in the validation set. All of the patients studied had tumors 5 cm or less in size and at least 15 axillary lymph nodes that had been examined. Data used for construction of the predictive models were available for all patients and included tumor size, number of nodes positive, patient age, quantitative estrogen receptor levels, quantitative progesterone receptor (PgR) levels, DNA flow cytometry-derived ploidy, and S-phase fraction. Logistic regression models were used to predict nodal status.

RESULTS

Multivariate predictive models were produced that used tumor size, patient age, S phase, and PgR as independent predictors. These models allowed identification of patient risks of being node positive ranging from 6%-79% and as having 10 or more positive nodes ranging from less than 1% to slightly more than 30%.

CONCLUSION

Addition of prognostic indicator information to tumor size can refine estimates of whether a patient is likely to be node positive. However, no patient subsets could be identified as having greater than 95% chance of being node negative or node positive.

IMPLICATIONS

These predictive models cannot alleviate the necessity of axillary node dissection for staging of breast cancer patients in situations in which nodal status would affect therapeutic decisions. Subsets of patients could be identified who had a less than 5% chance of having 10 or more positive nodes. Thus, some patients could be spared axillary dissection if it was being performed solely to identify patients with this high-risk feature.

Prognostic factors in early breast carcinoma

Several investigators, the SEER data, and the ECOG/Intergroup study have shown that patients with small tumors (< 0.5 cm) have a recurrence rate of less than 2%, compared to 20-25% for large tumors (> or = 5 cm). Nuclear grade and tumor differentiation are established indicators; however, the interobserver lack of concordance has thwarted their use in clinical trials. The presence of peritumoral lymphatic and blood vessel invasion (PLBI) is associated with a relative risk of recurrence of 4.7. The predictive value of the presence of hormone receptors in tumors is associated with a favorable disease free and overall survival difference of 8-10%; however, this advantage is being eroded by the early appearance of other factors, such as the epidermal growth factor receptor (EGFR), proliferative capacity (S-phase), nuclear grade, and HER-2/neu oncogene. Concordance among the different methods of hormone-receptor assay (immunocytochemical, sucrose gradient, and dextran-coated charcoal) is essential to refine the true value of these factors. DNA flow cytometry measurements of ploidy (DNA content) and S-phase fraction are the most characterized of the prognostic factors. There are conflicting reports regarding the clinical significance of ploidy status, while measurements of S-phase fraction clearly indicate a robust association with disease free and overall survival. Our data continue to show that S-phase, but not ploidy, can predict time to recurrence significantly in untreated patients, even when data are stratified for tumor size. HER-2/neu oncogene is expressed in about 50% of ductal carcinoma in situ and 14% of invasive ductal carcinoma. The presence of this oncogene at high copy number may be a useful independent marker of poor prognosis and may be associated with drug resistance and correlated with tumor recurrence and shorter survival. EGFR could be measured in most breast tumors, and the level of its expression has inversely correlated with estrogen receptor protein expression. The value of EGFR as a predictor of prognosis remains controversial and is still being investigated. Cathepsin-D provides a provocative biologic rationale but is hindered by different and incongruent methods of analysis. The majority of large studies with more than 3-years' follow-up suggests that high cathepsin-D levels may be predictive of greater recurrence and lower survival. Angiogenesis has been implicated as a critical component of the metastatic process. Early studies show that tumor angiogenesis is an independent and highly significant prognostic indicator, and its presence may suggest the selection of "anti-angiogenic therapy."(ABSTRACT TRUNCATED AT 400 WORDS)

Do American oncologists know how to use prognostic variables for patients with newly diagnosed primary breast cancer?

PURPOSE

This project was designed to investigate how American medical oncologists actually use prognostic information to treat primary breast cancer patients, and to study their difficulties in combining complex and sometimes contradictory information.

METHODS

A simple 2-page questionnaire was faxed in May and June 1993 to a sample of American medical oncologists who were members of the American Society of Clinical Oncology (ASCO).

RESULTS

When presented with simple case histories of patients with newly diagnosed invasive breast cancer and asked to assess prognosis on the basis of tumor size, number of involved axillary nodes, patient age, estrogen receptor level, and progesterone receptor level, there was a wide divergence of opinions about the probability of disease-free survival at 10 years (both for cases in which the patient received no adjuvant therapy and for those in which the patient did receive such therapy). The use of additional prognostic data (such as S-phase, tumor histologic and nuclear grading, and cathepsin D status) did not refine the estimates, but led to an equal or greater dispersion of estimates of prognosis.

CONCLUSION

There is a clear need for tools to help oncologists integrate prognostic information for primary breast cancer patients. Such tools might lead to greater accuracy and uniformity of prognostic estimates. Such tools might also help make clear what prognostic tests are worth using for routine clinical practice.

A technique for using neural network analysis to perform survival analysis of censored data

The purpose of this study was to demonstrate how a form of neural network analysis could be used to perform survival analysis on censored data, and to compare neural network analysis with the most commonly used technique for this type of analysis, Cox regression. In this study computer simulated data sets were used. The underlying rules connecting prognostic information to the hazard of death were defined to allow the construction of data sets with specific realistic properties that could be used to demonstrate situations in which neural network analysis had particular strengths in comparison with Cox regression modeling. Using these simulated data sets neural network analysis could produce successful predictive models, find interactions between variables, and recognize the importance of variables that contributed to the hazard rate as a complex function of the variables value and in situations where the proportionality of hazards assumption was violated. It was also demonstrated that neural network analysis was not a 'black box', but could lead to useful insights into the roles played by different prognostic variables in determining patient outcome.

Cathepsin D by western blotting and immunohistochemistry: failure to confirm correlations with prognosis in node-negative breast cancer

PURPOSE

We attempted to replicate and improve on our previous study (N Engl J Med 322:297-302, 1990) that showed that 34-kd cathepsin D levels as determined by Western blotting strongly correlated with disease-free survival (DFS) and overall survival (OS) of axillary node-negative (N-) breast cancer patients. We also examined the prognostic significance of cathepsin D measured by immunohistochemistry (IHC) in these patients.

PATIENTS AND METHODS

Western blotting was performed on cytosols from frozen tumor specimens of 927 N- breast cancer patients in the San Antonio Breast Tumor Bank. The monoclonal antibody M1G8 was used to detect cathepsin D (in previous study, a polyclonal antiserum had been used). The same monoclonal antibody was also used for frozen-section IHC staining of tumor specimens from 562 N- patients from the same tumor bank. Levels of cathepsin D expression were then correlated with DFS and OS.

RESULTS

Although the levels of cathepsin D expression as measured by Western blotting and IHC correlated with each other and with levels of cathepsin D measured in previous work using Western blotting with the polyclonal antiserum, in this present study, using the monoclonal antibody M1G8, we were unable to demonstrate that cathepsin D expression (measured by either Western blotting or by IHC) correlates with DFS or OS.

CONCLUSION

In this study, cathepsin D expression as determined by either Western blotting or IHC using the monoclonal antibody M1G8 failed to improve the prognostic evaluation of N- breast cancer patients. The role of cathepsin D expression as a prognostic factor is still not precisely defined, raising questions about its use in the routine evaluation of N- breast cancer patients.

A practical view of prognostic factors for staging, adjuvant treatment planning, and as baseline studies for possible future therapy

At this time the initial prognostic assessment of breast cancer patients is still most powerfully driven by basic histopathologic information, axillary nodal involvement, and tumor size. Estrogen and progesterone receptor status are important initial pieces of information for many patients, but this information is more important in deciding the most appropriate type of treatment, rather than the prognosis of the patient. Histologic and nuclear grading can provide important prognostic information, but broader application of this information awaits better methods to ensure accuracy and decrease intraobserver variability. Whether flow cytometry-derived information can be used to select patient subsets at very low risk of relapse awaits prospective validation in cooperative group trials. A number of new prognostic tests such as cathepsin D that have shown promise in some studies await definitive prospective validation. Further development of techniques to integrate prognostic factor information and the use of this information in individualized prognostic factor decisions is needed.

Case report: use of insulin-like growth factor-I gene expression to distinguish between breast and ovarian cancer

Autocrine expression of polypeptide growth factors may be important in the growth regulation of cancer cells. Different growth factor activities have been identified in a variety of tumors. This article describes a case of malignant ascites in a patient recently treated for breast cancer. The use of growth factor mRNA expression as a factor to differentiate between breast and ovarian origins of cancer cells contained in malignant ascites was examined. Expression of insulin-like growth factor-I (IGF-I), IGF-II, and transforming growth factor alpha mRNA was examined by ribonuclease protection assay. The tumor cells expressed IGF-II and transforming growth factor alpha, but not IGF-I mRNA. This pattern of growth factor expression is compatible with a breast cancer primary of the malignant cells contained in the ascites fluid. Therefore, IGF-I mRNA expression may be useful in distinguishing between adenocarcinomas of breast or ovarian origins.

Phase II trial of piroxantrone in metastatic breast cancer. A Southwest Oncology Group study

Thirty-two eligible patients with advanced metastatic breast cancer who had received no more than 1 prior chemotherapy regimen for metastatic disease (16 had received prior doxorubicin) were treated with piroxantrone at a dose of 120 mg/m2 intravenously every 21 days. In the twenty-seven patients evaluable for response, two partial responses were seen. Toxicities observed were primarily hematologic with grade 3 or greater granulocytopenia occurring in 34% of the patients. One patient developed symptomatic congestive heart failure at a total cumulative dose of 960 mg/m2. We conclude that piroxantrone given at this dose and schedule has minimal activity in patients with metastatic breast cancer.

Survival analysis of censored data: neural network analysis detection of complex interactions between variables

Neural networks can be used as pattern recognition systems in complex data sets. We are exploring their utility in performing survival analysis to predict time to relapse or death. This technique has the potential to find easily some types of very complex interactions in data that would not be easily recognized by conventional statistical methods. In this paper we demonstrate that there are several ways neural networks can be used to find three-way interactions among variables. Thus, in data sets where such complex interactions exist, neural networks may find utility in detecting such interactions and in helping to produce predictive models.

Prognostic factors: rationale and methods of analysis and integration

With the proliferation of potential prognostic factors for breast cancer, it is becoming increasingly more difficult for physicians and patients to integrate the information provided by these factors into a single accurate prediction of clinical outcome. Here we review Cox's proportional hazards model, recursive partitioning, correspondence analysis, and neural networks for their respective capabilities in analyzing censored survival data in the presence of multiple prognostic factors, and we present some clinical applications where these models have been used.

Failure of GM-CSF to permit dose-escalation in an every other week dose-intensive regimen for advanced breast cancer

BACKGROUND

In an attempt to improve dose intensity and therapeutic effectiveness in breast cancer, GM-CSF was incorporated into a multi-drug every other week chemotherapy regimen that had been previously reported to have promising activity, but whose dose-limiting toxicity was neutropenia.

PATIENTS AND METHODS

A Phase I-II study in patients with locally advanced or metastatic breast cancer was initiated using GM-CSF and a 5-drug chemotherapy regimen employing oral cyclophosphamide daily for 7 days and doxorubicin, vincristine, methotrexate, 5-fluorouracil, and leucovorin IV every 2 weeks for 10 courses. In the first 8 patients, GM-CSF in escalating doses (1-20 micrograms/kg s.c. per day) was given on days 8-13 of each 2 week cycle. In the last 12 patients, GM-CSF was given on days 3-14 of each cycle in an attempt to improve its effectiveness by prolonging treatment duration.

RESULTS

The regimen was poorly tolerated. Only 10 patients completed all 10 courses of treatment, and most of those required dose delays and/or reductions. GM-CSF failed to reduce neutropenia when given by either schedule. Furthermore, thrombocytopenia was severe and progressive, especially with the more prolonged GM-CSF schedule in which the mean lowest nadir platelet count was 15,000/microliters. Anemia, fatigue, mucositis, and neutropenic fevers were also common, and dose escalations were not possible in any patient. Central venous catheter complications were also common. Complete or partial remissions were observed in 15 of 20 patients, but response durations were brief.

CONCLUSIONS

GM-CSF in two different schedules failed to ameliorate myelosuppression when used in combination with this multiple drug, every other week regimen. Neutropenia and, especially, thrombocytopenia remained dose limiting.