Christos Hatzis

Measurement of Residual Breast Cancer Burden to Predict Survival After Neoadjuvant Chemotherapy

PURPOSE To measure residual disease after neoadjuvant chemotherapy in order to improve the prognostic information that can be obtained from evaluating pathologic response. PATIENTS AND METHODS Pathologic slides and reports were reviewed from 382 patients in two different treatment cohorts: sequential paclitaxel (T) then fluorouracil, doxorubicin, and cyclophosphamide (FAC) in 241 patients; and a single regimen of FAC in 141 patients. Residual cancer burden (RCB) was calculated as a continuous index combining pathologic measurements of primary tumor (size and cellularity) and nodal metastases (number and size) for prediction of distant relapse-free survival (DRFS) in multivariate Cox regression analyses. RESULTS RCB was independently prognostic in a multivariate model that included age, pretreatment clinical stage, hormone receptor status, hormone therapy, and pathologic response (pathologic complete response [pCR] v residual disease [RD]; hazard ratio = 2.50; 95% CI 1.70 to 3.69; P < .001). Minimal RD (RCB-I) in 17% of patients carried the same prognosis as pCR (RCB-0). Extensive RD (RCB-III) in 13% of patients was associated with poor prognosis, regardless of hormone receptor status, adjuvant hormone therapy, or pathologic American Joint Committee on Cancer stage of residual disease. The generalizability of RCB for prognosis of distant relapse was confirmed in the FAC-treated validation cohort. CONCLUSION RCB determined from routine pathologic materials represented the distribution of RD, was a significant predictor of DRFS, and can be used to define categories of near-complete response and chemotherapy resistance.

A Genomic Predictor of Response and Survival Following Taxane-Anthracycline Chemotherapy for Invasive Breast Cancer

CONTEXT Prediction of high probability of survival from standard cancer treatments is fundamental for individualized cancer treatment strategies. OBJECTIVE To develop a predictor of response and survival from chemotherapy for newly diagnosed invasive breast cancer. DESIGN, SETTING, AND PATIENTS Prospective multicenter study conducted from June 2000 to March 2010 at the M. D. Anderson Cancer Center to develop and test genomic predictors for neoadjuvant chemotherapy. Patients were those with newly diagnosed ERBB2 (HER2 or HER2/neu)-negative breast cancer treated with chemotherapy containing sequential taxane and anthracycline-based regimens (then endocrine therapy if estrogen receptor [ER]-positive). Different predictive signatures for resistance and response to preoperative (neoadjuvant) chemotherapy (stratified according to ER status) were developed from gene expression microarrays of newly diagnosed breast cancer (310 patients). Breast cancer treatment sensitivity was then predicted using the combination of signatures for (1) sensitivity to endocrine therapy, (2) chemoresistance, and (3) chemosensitivity, with independent validation (198 patients) and comparison with other reported genomic predictors of chemotherapy response. MAIN OUTCOME MEASURES Distant relapse-free survival (DRFS) if predicted treatment sensitive and absolute risk reduction ([ARR], difference in DRFS between 2 predicted groups) at median follow-up (3 years). RESULTS Patients in the independent validation cohort (99% clinical stage II-III) who were predicted to be treatment sensitive (28%) had 56% (95% CI, 31%-78%) probability of excellent pathologic response and DRFS of 92% (95% CI, 85%-100%), with an ARR of 18% (95% CI, 6%-28%). Survival was predicted in ER-positive (30% predicted sensitive; DRFS, 97% [95% CI, 91%-100%]; ARR, 11% [95% CI, 0.1%-21%]) and ER-negative (26% predicted sensitive; DRFS, 83% [95% CI, 68%-100%]; ARR, 26% [95% CI, 4%-48%]) subsets and was significant in multivariate analysis. Other genomic predictors showed paradoxically worse survival for patients predicted to be responsive to chemotherapy. CONCLUSION A genomic predictor combining ER status, predicted chemoresistance, predicted chemosensitivity, and predicted endocrine sensitivity identified patients with high probability of survival following taxane and anthracycline chemotherapy.

Commercialized Multigene Predictors of Clinical Outcome for Breast Cancer

In the past 5 years, a number of commercialized multigene prognostic and predictive tests have entered the complex and expanding landscape of breast cancer companion diagnostics. These tests have used a variety of formats ranging from the familiar slide-based assays of immunohistochemistry and fluorescence in situ hybridization to the nonmorphology-driven molecular platforms of quantitative multiplex real-time polymerase chain reaction and genomic microarray profiling. In this review, 14 multigene assays are evaluated as to their scientific validation, current clinical utility, regulatory approval status, and estimated cost-benefit ratio. Emphasis is placed on two tests: oncotype DX and MammaPrint. Current evidence indicates that the oncotype DX test has the advantages of earlier commercial launch, wide acceptance for payment by third-party payors in the U.S., ease of use of formalin-fixed paraffin-embedded tissues, recent listing by the American Society of Clinical Oncology Breast Cancer Tumor Markers Update Committee as recommended for use, continuous scoring system algorithm, ability to serve as both a prognostic test and predictive test for certain hormonal and chemotherapeutic agents, demonstrated cost-effectiveness in one published study, and a high accrual rate for the prospective validation clinical trial (Trial Assigning Individualized Options for Treatment). The MammaPrint assay has the advantages of a 510(k) clearance by the U.S. Food and Drug Administration, a larger gene number, which may enhance further utility, and a potentially wider patient eligibility, including lymph node-positive, estrogen receptor (ER)-negative, and younger patients being accrued into the prospective trial (Microarray in Node-Negative Disease May Avoid Chemotherapy). A number of other assays have specific predictive goals that are most often focused on the efficacy of tamoxifen in ER-positive patients, such as the two-gene ratio test and the cytochrome P450 CYP2D6 genotyping assay.

Genomic Index of Sensitivity to Endocrine Therapy for Breast Cancer

PURPOSE We hypothesize that measurement of gene expression related to estrogen receptor α (ER; gene name ESR1) within a breast cancer sample represents intrinsic tumoral sensitivity to adjuvant endocrine therapy. METHODS A genomic index for sensitivity to endocrine therapy (SET) index was defined from genes coexpressed with ESR1 in 437 microarray profiles from newly diagnosed breast cancer, unrelated to treatment or outcome. The association of SET index and ESR1 levels with distant relapse risk was evaluated from microarrays of ER-positive breast cancer in two cohorts who received 5 years of tamoxifen alone as adjuvant endocrine therapy (n = 225 and 298, respectively), a cohort who received neoadjuvant chemotherapy followed by tamoxifen and/or aromatase inhibition (n = 122), and two cohorts who received no adjuvant systemic therapy (n = 208 and 133, respectively). RESULTS The SET index (165 genes) was significantly associated with distant relapse or death risk in both tamoxifen-treated cohorts (hazard ratio [HR] = 0.70, 95% CI, 0.56 to 0.88, P = .002; and HR = 0.76, 95% CI, 0.63 to 0.93, P = .007) and in the chemo-endocrine-treated cohort (HR = 0.19; 95% CI, 0.05 to 0.69, P = .011) independently from pathologic response to chemotherapy, but was not prognostic in two untreated cohorts. No distant relapse or death was observed after tamoxifen alone if node-negative and high SET or after chemo-endocrine therapy if intermediate or high SET. CONCLUSION The SET index of ER-related transcription predicted survival benefit from adjuvant endocrine therapy, not inherent prognosis. Prior chemotherapy seemed to enhance the efficacy of adjuvant endocrine therapy related to SET index.

Genomic Grade Index Is Associated With Response to Chemotherapy in Patients With Breast Cancer

PURPOSE The genomic grade index (GGI) is a 97-gene measure of histological tumor grade. High GGI is associated with decreased relapse-free survival in patients receiving either endocrine or no systemic adjuvant therapy. Herein we examined whether GGI predicts pathologic response to neoadjuvant chemotherapy in patients with HER-2-normal breast cancer. METHODS Gene expression data (gene chips) was generated from fine-needle aspiration biopsies (n = 229) prospectively collected before neoadjuvant paclitaxel, fluorouracil, doxorubicin, and cyclophosphamide chemotherapy. Pathologic response was quantified using the residual cancer burden (RCB) method. The association between the GGI and pathologic response was assessed in univariate and multivariate analyses. The performance of a response predictor combining clinical variables and GGI was evaluated under cross-validation. Results Eighty-five percent of grade 1 tumors had low GGI, 89% of grade 3 tumors had high GGI, and 63% of grade 2 tumors had low GGI. Among both estrogen receptor (ER)-positive and -negative cancers, high GGI score was associated with pathologic complete response (RCB-0) or minimal residual disease (RCB-1). A multivariate model combining GGI and clinical parameters had an overall accuracy of 71%, compared with 58% for the GGI alone, for prediction of pathologic response. However, high GGI score was also associated with significantly worse distant relapse-free survival in patients with ER-positive cancer (P = .005), and was not associated with survival in patients with ER-negative cancer. CONCLUSION High GGI is associated with increased sensitivity to neoadjuvant paclitaxel plus fluorouracil, adriamycin, and cyclophosphamide chemotherapy in both ER-negative and ER-positive patients, but it remains a predictor of worse survival in ER-positive patients.

Molecular Anatomy of Breast Cancer Stroma and Its Prognostic Value in Estrogen Receptor–Positive and –Negative Cancers

PURPOSE The purpose of this study was to identify genes enriched in breast cancer stroma, assess the stromal gene expression differences between estrogen receptor (ER) -positive and -negative cancers, and separately determine their prognostic value in these two subtypes of breast cancers. METHODS We compared gene expression profiles of pairs of fine-needle (stroma-poor) and core-needle (stroma-rich) biopsies from 37 cancers to identify stroma-associated genes. We defined stromal metagenes and tested their prognostic values in 684 node-negative patients who received no systemic adjuvant therapy and 259 tamoxifen-treated patients. RESULTS We identified 293 probe sets overexpressed in core biopsies; these included five highly coexpressed gene clusters (metagenes) corresponding to immune functions and extracellular matrix components. These genes showed quantitative and qualitative differences between ER-positive and ER-negative cancers. A B-cell/plasma cell metagene showed strong prognostic value in ER-positive highly proliferative cancers, a lesser prognostic value in ER-negative cancers, and no prognostic value in ER-positive cancers with low proliferation. The hazard ratio for distant relapse in the lowest compared with the highest tertile of the pooled prognostic data set was 4.29 (95% CI, 2.04 to 9.01; P = .001) in ER-positive cancers and 3.34 (95% CI, 1.60 to 6.97; P = .001) in ER-negative cancers. This remained significant in multivariate analysis including routine variables and other genomic prognostic scores. As a result of quantitative differences in this metagene between ER-positive and ER-negative cancers, different thresholds apply in the two subgroups. Other stromal metagenes had inconsistent prognostic value. CONCLUSION Among ER-negative and ER-positive highly proliferative cancers, a subset of tumors with high expression of a B-cell/plasma cell metagene carries a favorable prognosis.

Estrogen Receptor (ER) mRNA and ER-Related Gene Expression in Breast Cancers That Are 1% to 10% ER-Positive by Immunohistochemistry

PURPOSE We examined borderline estrogen receptor (ER) -positive cancers, defined as having 1% to 10% positivity by immunohistochemistry (IHC), to determine whether they show the same global gene-expression pattern and high ESR1 mRNA expression as ER-positive cancers or if they are more similar to ER-negative cancers. PATIENTS AND METHODS ER status was determined by IHC in 465 primary breast cancers and with the Affymetrix U133A gene chip. We compared expressions of ESR1 mRNA and a 106 probe set ER-associated gene signature score between ER-negative (n = 183), 1% to 9% (n = 25), 10% (n = 6), and more than 10% (n = 251) ER-positive cancers. We also assessed the molecular class by using the PAM50 classifier and plotted survival by ER status. RESULTS Among the 1% to 9%, 10%, and more than 10% ER IHC-positive patients, 24%, 67%, and 92% were also positive by ESR1 mRNA expression. The average ESR1 expression was significantly higher in the ≥ 10% ER-positive cohorts compared with the 1% to 9% or ER-negative cohort. The average ER gene signature scores were similar for the ER-negative and 1% to 9% IHC-positive patients and were significantly lower than in ≥ 10% ER-positive patients. Among the 1% to 9% ER-positive patients, 8% were luminal B and 48% were basal-like; among the 10% ER-positive patients, 50% were luminal. The overall survival rate of 1% to 9% ER-positive patients with cancer was between those of patients in the ≥ 10% ER-positive and ER-negative groups. CONCLUSION A minority of the 1% to 9% IHC ER-positive tumors show molecular features similar to those of ER-positive, potentially endocrine-sensitive tumors, whereas most show ER-negative, basal-like molecular characteristics. The safest clinical approach may be to use both adjuvant endocrine therapy and chemotherapy in this rare subset of patients.

Microtubule-associated protein-tau is a bifunctional predictor of endocrine sensitivity and chemotherapy resistance in estrogen receptor-positive breast cancer

Purpose: The clinical outcome for patients with breast cancer is influenced by the metastatic competence of the cancer and its sensitivity to endocrine therapy and chemotherapy. A molecular marker may be prognostic of outcome or predictive of response to therapy, or a combination of both. Experimental Design: We examined separately the prognostic and predictive values of tau mRNA expression in estrogen receptor (ER)–positive primary breast cancers in three patient cohorts. We used gene expression data from 209 untreated patients to assess the pure prognostic value of tau, data from 267 patients treated with adjuvant tamoxifen to assess predictive value for endocrine therapy, and data from 82 patients treated with preoperative paclitaxel followed by 5-fluorouracil, doxorubicin, and cyclophosphamide (paclitaxel/FAC) to assess predictive value for chemotherapy response. Wilcoxon rank sum test was used to compare tau expression between different outcome groups. Results: Higher tau mRNA expression showed borderline nonsignificant association with better prognosis in the absence of systemic adjuvant therapy. Higher tau mRNA expression was significantly associated with no recurrence (at 5 and 10 years, P = 0.005 and P = 0.05, respectively) in patients treated with tamoxifen, indicating a predictive value for endocrine therapy. Tau expression was significantly lower in patients who achieved pathologic complete response to paclitaxel/FAC chemotherapy (P < 0.001). Conclusion: This study suggests that high tau mRNA expression in ER-positive breast cancer indicates an endocrine-sensitive but chemotherapy-resistant disease. In contrast, low tau expression identifies a subset of ER-positive cancers that have poor prognosis with tamoxifen alone and may benefit from taxane-containing chemotherapy.

Effect of training-sample size and classification difficulty on the accuracy of genomic predictors

IntroductionAs part of the MicroArray Quality Control (MAQC)-II project, this analysis examines how the choice of univariate feature-selection methods and classification algorithms may influence the performance of genomic predictors under varying degrees of prediction difficulty represented by three clinically relevant endpoints.MethodsWe used gene-expression data from 230 breast cancers (grouped into training and independent validation sets), and we examined 40 predictors (five univariate feature-selection methods combined with eight different classifiers) for each of the three endpoints. Their classification performance was estimated on the training set by using two different resampling methods and compared with the accuracy observed in the independent validation set.ResultsA ranking of the three classification problems was obtained, and the performance of 120 models was estimated and assessed on an independent validation set. The bootstrapping estimates were closer to the validation performance than were the cross-validation estimates. The required sample size for each endpoint was estimated, and both gene-level and pathway-level analyses were performed on the obtained models.ConclusionsWe showed that genomic predictor accuracy is determined largely by an interplay between sample size and classification difficulty. Variations on univariate feature-selection methods and choice of classification algorithm have only a modest impact on predictor performance, and several statistically equally good predictors can be developed for any given classification problem.

KDD Cup 2001 report

This paper presents results and lessons from KDD Cup 2001. KDD Cup 2001 focused on mining biological databases. It involved three cutting-edge tasks related to drug design and genomics.