Marco Bonetti

Burdens and Benefits of Adjuvant Cyclophosphamide, Methotrexate, and Fluorouracil and Tamoxifen for Elderly Patients With Breast Cancer: The International Breast Cancer Study Group Trial VII

PURPOSE Information on the tolerability and efficacy of adjuvant chemoendocrine therapy for older women is limited. We studied these issues using the data collected as part of the International Breast Cancer Study Group Trial VII. PATIENTS AND METHODS Postmenopausal women with operable, node-positive breast cancer were randomized to receive either tamoxifen alone for 5 years (306 patients) or tamoxifen plus three consecutive cycles of classical cyclophosphamide (100 mg/m(2) orally days 1 to 14), methotrexate (40 mg/m(2) intravenous days 1 and 8), and fluorouracil (600 mg/m(2) intravenous days 1 and 8) every 28 days (CMF; 302 patients). The median follow-up was 8.0 years. RESULTS Among the 299 patients who received at least one dose of CMF, women 65 years of age or older (n = 76) had higher grades of toxicity compared with women less than 65 years old (n = 223) (P =.004). More women in the older age group compared with the younger women experienced grade 3 toxicity of any type (17% v 7%, respectively), grade 3 hematologic toxicity (9% v 5%, respectively), and grade 3 mucosal toxicity (4% v 1%, respectively). Older patients also received less than their expected CMF dose compared with younger postmenopausal women (P =.0008). The subjective burdens of treatment, however, were similar for younger and older patients based on quality-of-life measures (performance status, coping, physical well-being, mood, and appetite). For older patients, the 5-year disease-free survival (DFS) rates were 63% for CMF plus tamoxifen and 61% for tamoxifen alone (hazards ratio [HR], 1.00; 95% confidence interval [CI], 0.65 to 1.52; P =.99). For younger patients, the corresponding 5-year DFS rates were 61% and 53% (HR, 0.70; 95% CI, 0.53 to 0.91; P =.008), but the test for heterogeneity of CMF effect according to age group was not statistically significant. The reduced effectiveness of CMF among older women could not be attributed to dose reductions according to dose received. CONCLUSION CMF tolerability and effectiveness were both reduced for older patients compared with younger postmenopausal node-positive breast cancer patients who received tamoxifen for 5 years. The development and evaluation of less toxic and more effective chemotherapy regimens are required for high-risk elderly patients.

Early Start of Adjuvant Chemotherapy May Improve Treatment Outcome for Premenopausal Breast Cancer Patients With Tumors not Expressing Estrogen Receptors

PURPOSE The proper time to commence adjuvant chemotherapy after primary surgery for breast cancer is unknown. An analysis of the International (Ludwig) Breast Cancer Study Group (IBCSG) Trial V at a median follow-up of 11 years suggested that early initiation of adjuvant chemotherapy might improve outcome for premenopausal, node-positive patients whose tumors did not express any estrogen receptor (ER). PATIENTS AND METHODS We investigated the relationship between early initiation of adjuvant chemotherapy, ER status, and prognosis in 1,788 premenopausal, node-positive patients treated on IBCSG trials I, II, and VI. The disease-free survival for 599 patients (84 with ER-absent tumors) who commenced adjuvant chemotherapy within 20 days (early initiation) was compared with the disease-free survival for 1,189 patients (142 with ER-absent tumors) who started chemotherapy 21 to 86 days after surgery (conventional initiation). The median follow-up was 7.7 years. RESULTS Among patients with ER-absent tumors, the 10-year disease-free survival was 60% for the early initiation group compared with 34% for the conventional initiation group (226 patients; hazard ratio [HR], 0. 49; 95% confidence interval [CI], 0.33 to 0.72; P =.0003). This difference remained statistically significant in a Cox multiple regression analysis controlling for study group, number of positive nodes, tumor size, age, vessel invasion, and institution (HR, 0.60; 95% CI, 0.39 to 0.92; P =.019). Conversely, early initiation of chemotherapy did not significantly improve disease-free survival for patients with tumors expressing ER (1,562 patients; multiple regression HR, 0.93; 95% CI, 0.79 to 1.10; P =.40). CONCLUSION In premenopausal patients with ER-absent tumors, early initiation of systemic chemotherapy after primary surgery might improve outcome. Further confirmatory studies are required before any widespread modification of current clinical practice. In premenopausal patients with tumors expressing some ER, gains from early initiation are unlikely to be clinically significant.

A graphical method to assess treatment–covariate interactions using the Cox model on subsets of the data

We introduce the subpopulation treatment effect pattern plot (STEPP) method, designed to facilitate the interpretation of estimates of treatment effect derived from different but potentially overlapping subsets of clinical trial data. In particular, we consider sequences of subpopulations defined with respect to a covariate, and obtain confidence bands for the collection of treatment effects (here obtained from the Cox proportional hazards model) associated with the sequences. The method is aimed at determining whether the magnitude of the treatment effect changes as a function of the values of the covariate. We apply STEPP to a breast cancer clinical trial data set to evaluate the treatment effect as a function of the oestrogen receptor content of the primary tumour.

A nonparametric scoring algorithm for identifying informative genes from microarray data.

Microarray data routinely contain gene expression levels of thousands of genes. In the context of medical diagnostics, an important problem is to find the genes that are correlated with given phenotypes. These genes may reveal insights to biological processes and may be used to predict the phenotypes of new samples. In most cases, while the gene expression levels are available for a large number of genes, only a small fraction of these genes may be informative in classification with statistical significance. We introduce a nonparametric scoring algorithm that assigns a score to each gene based on samples with known classes. Based on these scores, we can find a small set of genes which are informative of their class, and subsequent analysis can be carried out with this set. This procedure is robust to outliers and different normalization schemes, and immediately reduces the size of the data with little loss of information. We study the properties of this algorithm and apply it to the data set from cancer patients. We quantify the information in a given set of genes by comparing its distribution of the score statistics to a set of distributions generated by permutations that preserve the correlation structure among the genes.

Identifying Breast Cancer Patients at High Risk for Bone Metastases

PURPOSE To identify patient populations at high risk for bone metastases at any time after diagnosis of operable breast cancer, because these patients are potential beneficiaries of treatment with bisphosphonates. PATIENTS AND METHODS We evaluated data from 6,792 patients who were randomized in International Breast Cancer Study Group clinical trials between 1978 and 1993. Median follow-up was 10. 7 years. A total of 1,275 patients (18.7%) presented with node-negative disease, whereas 3,354 patients (49.4%) had one to three and 2,163 patients (31.9%) had four or more involved axillary lymph nodes. We also assessed the incidence of subsequent bone metastases in the cohort of 1,220 patients who had a first event in local or regional sites or soft tissue alone. Median follow-up for this cohort was 7.7 years from first recurrence. RESULTS For the entire population with operable disease, the cumulative incidence of bone metastases at any time was 8.2% at 2 years from randomization and 27.3% at 10 years. The highest cumulative incidences of bone metastases at any time were among patients who had four or more involved axillary nodes at the time of diagnosis (14.9% at 2 years and 40.8% at 10 years) and among patients who had as their first event a local or regional recurrence or a recurrence in soft tissue, without any other overt metastases (21.1% at 2 years from first recurrence and 36.7% at 10 years). CONCLUSION Treatments to prevent bone metastases may have a major impact on the course of breast cancer and may be most efficiently studied in populations with several involved axillary nodes at the time of presentation and in populations with local or regional recurrence or recurrence in soft tissue.

Evaluation of Treatment-Effect Heterogeneity Using Biomarkers Measured on a Continuous Scale: Subpopulation Treatment Effect Pattern Plot

The discovery of biomarkers that predict treatment effectiveness has great potential for improving medical care, particularly in oncology. These biomarkers are increasingly reported on a continuous scale, allowing investigators to explore how treatment efficacy varies as the biomarker values continuously increase, as opposed to using arbitrary categories of expression levels resulting in a loss of information. In the age of biomarkers as continuous predictors (eg, expression level percentage rather than positive v negative), alternatives to such dichotomized analyses are needed. The purpose of this article is to provide an overview of an intuitive statistical approach-the subpopulation treatment effect pattern plot (STEPP)-for evaluating treatment-effect heterogeneity when a biomarker is measured on a continuous scale. STEPP graphically explores the patterns of treatment effect across overlapping intervals of the biomarker values. As an example, STEPP methodology is used to explore patterns of treatment effect for varying levels of the biomarker Ki-67 in the BIG (Breast International Group) 1-98 randomized clinical trial comparing letrozole with tamoxifen as adjuvant therapy for postmenopausal women with hormone receptor-positive breast cancer. STEPP analyses showed patients with higher Ki-67 values who were assigned to receive tamoxifen had the poorest prognosis and may benefit most from letrozole.

Duration of adjuvant chemotherapy for breast cancer: a joint analysis of two randomised trials investigating three versus six courses of CMF.

Cyclophosphamide, methotrexate and fluorouracil adjuvant combination chemotherapy for breast cancer is currently used for the duration of six monthly courses. We performed a joint analysis of two studies on the duration of adjuvant cyclophosphamide, methotrexate and fluorouracil in patients with node-positive breast cancer to investigate whether three courses of cyclophosphamide, methotrexate and fluorouracil might suffice. The International Breast Cancer Study Group Trial VI randomly assigned 735 pre- and perimenopausal patients to receive ‘classical’ cyclophosphamide, methotrexate and fluorouracil for three consecutive cycles, or the same chemotherapy for six consecutive cycles. The German Breast Cancer Study Group randomised 289 patients to receive either three or six cycles of i.v. cyclophosphamide, methotrexate and fluorouracil day 1, 8. Treatment effects were estimated using Cox regression analysis stratified by clinical trial without further adjustment for covariates. The 5-year disease-free survival per cents (±s.e.) were 54±2% for three cycles and 55±2% for six cycles (n=1024; risk ratio (risk ratio: CMF × 3/CMF × 6), 1.00; 95% confidence interval, 0.85 to 1.18; P=0.99). Use of three rather than six cycles was demonstrated to be adequate in both studies for patients at least 40-years-old with oestrogen-receptor-positive tumours (n=594; risk ratio, 0.86; 95% confidence interval, 0.68 to 1.08; P=0.19). In fact, results slightly favoured three cycles over six for this subgroup, and the 95% confidence interval excluded an adverse effect of more than 2% with respect to absolute 5-year survival. In contrast, three cycles appeared to be possibly inferior to six cycles for women less than 40-years-old (n=190; risk ratio, 1.25; 95% confidence interval, 0.87 to 1.80; P=0.22) and for women with oestrogen-receptor-negative tumours (n=302; risk ratio, 1.15; 95% confidence interval, 0.85 to 1.57; P=0.37). Thus, three initial cycles of adjuvant cyclophosphamide, methotrexate and fluorouracil chemotherapy were as effective as six cycles for older patients (40-years-old) with oestrogen-receptor-positive tumours, while six cycles of adjuvant cyclophosphamide, methotrexate and fluorouracil might still be required for other cohorts. Because endocrine therapy with tamoxifen and GnRH analogues is now available for younger women with oestrogen-receptor-positive tumours, the need for six cycles of cyclophosphamide, methotrexate and fluorouracil is unclear and requires further investigation.

A small sample study of the STEPP approach to assessing treatment-covariate interactions in survival data

A new, intuitive method has recently been proposed to explore treatment-covariate interactions in survival data arising from two treatment arms of a clinical trial. The method is based on constructing overlapping subpopulations of patients with respect to one (or more) covariates of interest and in observing the pattern of the treatment effects estimated across the subpopulations. A plot of these treatment effects is called a subpopulation treatment effect pattern plot. Here, we explore the small sample characteristics of the asymptotic results associated with the method and develop an alternative permutation distribution-based approach to inference that should be preferred for smaller sample sizes. We then describe an extension of the method to the case in which the pattern of estimated quantiles of survivor functions is of interest.

Power comparisons for an improved disease clustering test

The current note presents the power comparisons for disease clustering tests, as originally reported by Kulldorff et al. (Comput. Statist. Data Anal. 42 (2003) 665). A minor improvement to the implementation of the M-statistic, motivated by that work, results in dramatically higher power to detect clusters of disease.

Real time spatial cluster detection using interpoint distances among precise patient locations

BackgroundPublic health departments in the United States are beginning to gain timely access to health data, often as soon as one day after a visit to a health care facility. Consequently, new approaches to outbreak surveillance are being developed. When cases cluster geographically, an analysis of their spatial distribution can facilitate outbreak detection. Our method focuses on detecting perturbations in the distribution of pair-wise distances among all patients in a geographical region. Barring outbreaks, this distribution can be quite stable over time. We sought to exemplify the method by measuring its cluster detection performance, and to determine factors affecting sensitivity to spatial clustering among patients presenting to hospital emergency departments with respiratory syndromes.MethodsThe approach was to (1) define a baseline spatial distribution of home addresses for a population of patients visiting an emergency department with respiratory syndromes using historical data; (2) develop a controlled feature set simulation by inserting simulated outbreak data with varied parameters into authentic background noise, thereby creating semisynthetic data; (3) compare the observed with the expected spatial distribution; (4) establish the relative value of different alarm strategies so as to maximize sensitivity for the detection of clustering; and (5) measure factors which have an impact on sensitivity.ResultsOverall sensitivity to detect spatial clustering was 62%. This contrasts with an overall alarm rate of less than 5% for the same number of extra visits when the extra visits were not characterized by geographic clustering. Clusters that produced the least number of alarms were those that were small in size (10 extra visits in a week, where visits per week ranged from 120 to 472), diffusely distributed over an area with a 3 km radius, and located close to the hospital (5 km) in a region most densely populated with patients to this hospital. Near perfect alarm rates were found for clusters that varied on the opposite extremes of these parameters (40 extra visits, within a 250 meter radius, 50 km from the hospital).ConclusionMeasuring perturbations in the interpoint distance distribution is a sensitive method for detecting spatial clustering. When cases are clustered geographically, there is clearly power to detect clustering when the spatial distribution is represented by the M statistic, even when clusters are small in size. By varying independent parameters of simulated outbreaks, we have demonstrated empirically the limits of detection of different types of outbreaks.