Breast Cancer
Conditions
Keywords
stage II breast cancer, stage IIIA breast cancer, stage IIIB breast cancer, stage IIIC breast cancer, stage IV breast cancer, HER2-negative breast cancer, HER2-positive breast cancer
Brief summary
RATIONALE: Imaging procedures, such as diffusion-weighted magnetic resonance imaging (DWI) and dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI), may help in evaluating how well patients with breast cancer respond to treatment. PURPOSE: This research trial studies DWI and DCE-MRI in assessing treatment response in patients with breast cancer undergoing neoadjuvant chemotherapy.
Detailed description
OBJECTIVES: Primary * To determine if the change in tumor apparent diffusion coefficient (ADC) value measured from each treatment timepoint to baseline is predictive of pathologic complete response (pCR). Secondary * To determine if the combined measurement of change in tumor ADC value, change in tumor volume, and change in peak signal-enhancement ratio (SER) is predictive of pCR. * To investigate the relative effectiveness of the individual measurements, change in tumor ADC value, change in tumor volume, and change in peak SER for predicting pCR in experimental treatment arms. * To assess the test-retest reproducibility of ADC metrics applied to breast tumors. OUTLINE: This is a multicenter study. Patients undergo diffusion-weighted magnetic resonance imaging (DWI) at baseline, after week 3 of neoadjuvant paclitaxel regimen, and prior to and after completion of 4 courses of neoadjuvant chemotherapy. Patients then undergo surgery. Patients undergo DWI prior to contrast administration for dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI). After completion of treatment procedure, patients are followed up for 5 years on the I-SPY 2 TRIAL.
Interventions
PROCEDUREdiffusion-weighted magnetic resonance imaging
diffusion-weighted magnetic resonance imaging examination and subsequent radiologist interpretation
Sponsors
National Cancer Institute (NCI)
American College of Radiology Imaging Network
Study design
Allocation
NA
Intervention model
SINGLE_GROUP
Primary purpose
DIAGNOSTIC
Masking
NONE
Eligibility
Sex/Gender
FEMALE
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
DISEASE CHARACTERISTICS: * Meets I-SPY 2 TRIAL inclusion criteria * High-risk for recurrent disease PATIENT CHARACTERISTICS: * Able to tolerate imaging required by protocol PRIOR CONCURRENT THERAPY: * Not specified
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Pathologic Complete Response (pCR) | Surgery | Pathologic complete response (pCR) is defined as the lack of all signs of cancer in tissue samples removed during surgery after Neoadjuvant treatment for Breast cancer. ie., no residual invasive disease in either breast or axillary lymph nodes after neoadjuvant therapy (ypT0/is, ypN0) Histopathologic analysis was performed using the Residual Cancer Burden system |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Determine the Accuracy of Predictive Models Including Covariates for Combined Measurement of Change in Tumor ADC Value, Change in Tumor Volume, and Other Variables | baseline and mid-treatment | Accuracy will be measured as the Area under the Receiver Operating Characteristic Curve (AUC) Predictive logistic regression modeling was performed in 207 patients with complete mid-treatment ΔADC and ΔFTV data. To build prediction models with ADC and other variables, a data-splitting approach was used where a randomly selected 60% of participants (124 patients), stratified according to pCR status and tumor subtype, were selected as the training data set and the rest (86 patients) as the test set. Logistic regression with backward variable selection was used to construct the prediction models, which were then applied to the remaining 40% of the data to obtain predictive scores for each participant. |
| Repeatability Coefficient (RC)Test-retest Metric for Reproducibility of ADC as Applied to Breast Tumors | baseline (pre-treatment) or after 3 weeks of taxane-based treatment (early-treatment) | within-subject standard deviation (wSD) Repeatability coefficient (RC): \[RC = 2.77\*wSD\] (units: 10E-3 mm/sec\^2) Smaller values of RC, bounded \[0, ...), represent agreement |
| Functional Tumor Volume (FTV) as a Predictor of Pathologic Complete Response (pCR) | Surgery | Pathologic complete response (pCR) is defined as the lack of all signs of cancer in tissue samples removed during surgery after Neoadjuvant treatment for Breast cancer. ie., no residual invasive disease in either breast or axillary lymph nodes after neoadjuvant therapy (ypT0/is, ypN0) Histopathologic analysis was performed using the Residual Cancer Burden system Functional tumor volume (FTV) (units cm3) was computed by summing all tumor voxels meeting specific enhancement criteria, with customized thresholds for each site to account for variability in MR imaging systems |
| ICC Test-retest Metric for Reproducibility of ADC as Applied to Breast Tumors | baseline (pre-treatment) or after 3 weeks of taxane-based treatment (early-treatment) | Test and retest DWI measurements for a given patient were performed on the same day in a single imaging session. Intraclass correlation coefficient (ICC) is derived from the analysis of variance (ANOVA) model estimates (Barnhart,Haber, Lin 2007), Larger values of ICC (bounded \[-1,1\]) represent agreement |
| Agreement Index (AI) Test-retest Metric for Reproducibility of ADC as Applied to Breast Tumors | baseline (pre-treatment) or after 3 weeks of taxane-based treatment (early-treatment) | Test and retest DWI measurements for a given patient were performed on the same day in a single imaging session. Agreement index (AI): (Zhang, Wang, Duan - 2014) is based on the data's overall ranking. AI confidence intervals were obtained via bootstrap method Larger values AI (bounded \[0.5,1\]) represent agreement |
| Within-subject Coefficient of Variation (wCV) Test-retest Metric for Reproducibility of ADC as Applied to Breast Tumors | baseline (pre-treatment) or after 3 weeks of taxane-based treatment (early-treatment) | within-subject standard deviation (wSD) Within-subject coefficient of variation (wCV): \[wCV = 100%\*wSD/mean\] Smaller values of wCV bounded for \[0,...) represent better agreement |
Countries
United States
Participant flow
Participants by arm
| Arm | Count |
|---|---|
| Diffusion Weighted-MRI Participants on all arms of the I-SPY II trial with both a diffusion-weighted magnetic resonance imaging (DWI-MRI) scan at baseline and 1 post-baseline timepoint (early-treatment, mid-treatment, or post-treatment). The experimental component/intervention is whether DW-MRI can predict therapeutic response in women receiving neoadjuvant treatment for breast cancer. | 242 |
| Total | 242 |
Withdrawals & dropouts
| Period | Reason | FG000 |
|---|---|---|
| Overall Study | Baseline Imaging failed QC requirements | 9 |
| Overall Study | Ineligible | 18 |
| Overall Study | No Acceptable post-baseline imaging | 21 |
| Overall Study | Not Randomized in Parent study | 116 |
Baseline characteristics
| Characteristic | Diffusion Weighted-MRI |
|---|---|
| Age, Continuous | 48.1 years STANDARD_DEVIATION 10.4 |
| Ethnicity (NIH/OMB) Hispanic or Latino | 23 Participants |
| Ethnicity (NIH/OMB) Not Hispanic or Latino | 154 Participants |
| Ethnicity (NIH/OMB) Unknown or Not Reported | 65 Participants |
| Race (NIH/OMB) American Indian or Alaska Native | 0 Participants |
| Race (NIH/OMB) Asian | 16 Participants |
| Race (NIH/OMB) Black or African American | 26 Participants |
| Race (NIH/OMB) More than one race | 0 Participants |
| Race (NIH/OMB) Native Hawaiian or Other Pacific Islander | 1 Participants |
| Race (NIH/OMB) Unknown or Not Reported | 26 Participants |
| Race (NIH/OMB) White | 173 Participants |
| Sex/Gender, Customized Female | 242 Participants |
Adverse events
| Event type | EG000 affected / at risk |
|---|---|
| deaths Total, all-cause mortality | 0 / 406 |
| other Total, other adverse events | 0 / 406 |
| serious Total, serious adverse events | 0 / 406 |
Outcome results
Primary
Pathologic Complete Response (pCR)
Pathologic complete response (pCR) is defined as the lack of all signs of cancer in tissue samples removed during surgery after Neoadjuvant treatment for Breast cancer. ie., no residual invasive disease in either breast or axillary lymph nodes after neoadjuvant therapy (ypT0/is, ypN0) Histopathologic analysis was performed using the Residual Cancer Burden system
Time frame: Surgery
| Arm | Measure | Group | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|---|
| Early Treatment Change | Pathologic Complete Response (pCR) | Pathological Complete Responders (pCR) | 71 Participants |
| Early Treatment Change | Pathologic Complete Response (pCR) | Non-Responders(pCR-) | 156 Participants |
| Mid-Treatment Change | Pathologic Complete Response (pCR) | Pathological Complete Responders (pCR) | 70 Participants |
| Mid-Treatment Change | Pathologic Complete Response (pCR) | Non-Responders(pCR-) | 140 Participants |
| Post-Treatment Change | Pathologic Complete Response (pCR) | Pathological Complete Responders (pCR) | 63 Participants |
| Post-Treatment Change | Pathologic Complete Response (pCR) | Non-Responders(pCR-) | 123 Participants |
Comparison: Receiver Operating Characteristic curves and the corresponding Area Under the Curve (AUC), were estimated for the change in Apparent Diffusion Coefficients (ADC1 - ADC0)/ADC0 (Test: %change in ADC; Reference: pCR) detect a difference of 0.15 between the AUC under H0: AUC=0.5 and an AUC under the alternative hypothesis of 0.65. It was assumed that the number of pCR non-responders would be approximately 2.7 times greater than the number of complete respondersp-value: 0.484Z-test
Comparison: Receiver Operating Characteristic curves and the corresponding Area Under the Curve (AUC), were estimated for the change in Apparent Diffusion Coefficients (ADC2 - ADC0)/ADC0 (Test: %change in ADC; Reference: pCR) detect a difference of 0.15 between the AUC under H0: AUC=0.5 and an AUC under the alternative hypothesis of 0.65. It was assumed that the number of pCR non-responders would be approximately 2.7 times greater than the number of complete respondersp-value: 0.017Z-test
Comparison: Receiver Operating Characteristic curves and the corresponding Area Under the Curve (AUC), were estimated for the change in Apparent Diffusion Coefficients (ADC3 - ADC0)/ADC0 (Test: %change in ADC; Reference: pCR) detect a difference of 0.15 between the AUC under H0: AUC=0.5 and an AUC under the alternative hypothesis of 0.65. It was assumed that the number of pCR non-responders would be approximately 2.7 times greater than the number of complete respondersp-value: 0.013Z-test
Secondary
Agreement Index (AI) Test-retest Metric for Reproducibility of ADC as Applied to Breast Tumors
Test and retest DWI measurements for a given patient were performed on the same day in a single imaging session. Agreement index (AI): (Zhang, Wang, Duan - 2014) is based on the data's overall ranking. AI confidence intervals were obtained via bootstrap method Larger values AI (bounded \[0.5,1\]) represent agreement
Time frame: baseline (pre-treatment) or after 3 weeks of taxane-based treatment (early-treatment)
| Arm | Measure | Value (NUMBER) |
|---|---|---|
| Early Treatment Change | Agreement Index (AI) Test-retest Metric for Reproducibility of ADC as Applied to Breast Tumors | 0.83 probability |
Secondary
Determine the Accuracy of Predictive Models Including Covariates for Combined Measurement of Change in Tumor ADC Value, Change in Tumor Volume, and Other Variables
Accuracy will be measured as the Area under the Receiver Operating Characteristic Curve (AUC) Predictive logistic regression modeling was performed in 207 patients with complete mid-treatment ΔADC and ΔFTV data. To build prediction models with ADC and other variables, a data-splitting approach was used where a randomly selected 60% of participants (124 patients), stratified according to pCR status and tumor subtype, were selected as the training data set and the rest (86 patients) as the test set. Logistic regression with backward variable selection was used to construct the prediction models, which were then applied to the remaining 40% of the data to obtain predictive scores for each participant.
Time frame: baseline and mid-treatment
| Arm | Measure | Value (NUMBER) |
|---|---|---|
| Early Treatment Change | Determine the Accuracy of Predictive Models Including Covariates for Combined Measurement of Change in Tumor ADC Value, Change in Tumor Volume, and Other Variables | 0.71 probability |
| Mid-Treatment Change | Determine the Accuracy of Predictive Models Including Covariates for Combined Measurement of Change in Tumor ADC Value, Change in Tumor Volume, and Other Variables | 0.72 probability |
| Post-Treatment Change | Determine the Accuracy of Predictive Models Including Covariates for Combined Measurement of Change in Tumor ADC Value, Change in Tumor Volume, and Other Variables | 0.57 probability |
Comparison: AUC (optimized) = AUC (ΔADC)p-value: 0.032z-test
Secondary
Functional Tumor Volume (FTV) as a Predictor of Pathologic Complete Response (pCR)
Pathologic complete response (pCR) is defined as the lack of all signs of cancer in tissue samples removed during surgery after Neoadjuvant treatment for Breast cancer. ie., no residual invasive disease in either breast or axillary lymph nodes after neoadjuvant therapy (ypT0/is, ypN0) Histopathologic analysis was performed using the Residual Cancer Burden system Functional tumor volume (FTV) (units cm3) was computed by summing all tumor voxels meeting specific enhancement criteria, with customized thresholds for each site to account for variability in MR imaging systems
Time frame: Surgery
| Arm | Measure | Group | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|---|
| Early Treatment Change | Functional Tumor Volume (FTV) as a Predictor of Pathologic Complete Response (pCR) | Pathological Complete Responders (pCR) | 71 Participants |
| Early Treatment Change | Functional Tumor Volume (FTV) as a Predictor of Pathologic Complete Response (pCR) | Non-Responders(pCR-) | 156 Participants |
| Mid-Treatment Change | Functional Tumor Volume (FTV) as a Predictor of Pathologic Complete Response (pCR) | Pathological Complete Responders (pCR) | 70 Participants |
| Mid-Treatment Change | Functional Tumor Volume (FTV) as a Predictor of Pathologic Complete Response (pCR) | Non-Responders(pCR-) | 140 Participants |
| Post-Treatment Change | Functional Tumor Volume (FTV) as a Predictor of Pathologic Complete Response (pCR) | Pathological Complete Responders (pCR) | 63 Participants |
| Post-Treatment Change | Functional Tumor Volume (FTV) as a Predictor of Pathologic Complete Response (pCR) | Non-Responders(pCR-) | 123 Participants |
Comparison: Receiver Operating Characteristic curves and the corresponding Area Under the Curve (AUC), were estimated for the change in functional tumor volumes (FTV1 - FTV0)/FTV0 (Test: %change in FTV; Reference: pCR) detect a difference of 0.15 between the AUC under H0: AUC=0.5 and an AUC under the alternative hypothesis of 0.65. It was assumed that the number of pCR non-responders would be approximately 2.7 times greater than the number of complete respondersp-value: <0.001Z-test
Comparison: Receiver Operating Characteristic curves and the corresponding Area Under the Curve (AUC), were estimated for the change in functional tumor volumes (FTV2 - FTV0)/FTV0 (Test: %change in FTV; Reference: pCR) detect a difference of 0.15 between the AUC under H0: AUC=0.5 and an AUC under the alternative hypothesis of 0.65. It was assumed that the number of pCR non-responders would be approximately 2.7 times greater than the number of complete respondersp-value: <0.001Z-test
Comparison: Receiver Operating Characteristic curves and the corresponding Area Under the Curve (AUC), were estimated for the change in functional tumor volumes (FTV3 - FTV0)/FTV0 (Test: %change in FTV; Reference: pCR) detect a difference of 0.15 between the AUC under H0: AUC=0.5 and an AUC under the alternative hypothesis of 0.65. It was assumed that the number of pCR non-responders would be approximately 2.7 times greater than the number of complete respondersp-value: <0.001Z-test
Secondary
ICC Test-retest Metric for Reproducibility of ADC as Applied to Breast Tumors
Test and retest DWI measurements for a given patient were performed on the same day in a single imaging session. Intraclass correlation coefficient (ICC) is derived from the analysis of variance (ANOVA) model estimates (Barnhart,Haber, Lin 2007), Larger values of ICC (bounded \[-1,1\]) represent agreement
Time frame: baseline (pre-treatment) or after 3 weeks of taxane-based treatment (early-treatment)
| Arm | Measure | Value (NUMBER) |
|---|---|---|
| Early Treatment Change | ICC Test-retest Metric for Reproducibility of ADC as Applied to Breast Tumors | 0.97 correlation coefficient |
Secondary
Repeatability Coefficient (RC)Test-retest Metric for Reproducibility of ADC as Applied to Breast Tumors
within-subject standard deviation (wSD) Repeatability coefficient (RC): \[RC = 2.77\*wSD\] (units: 10E-3 mm/sec\^2) Smaller values of RC, bounded \[0, ...), represent agreement
Time frame: baseline (pre-treatment) or after 3 weeks of taxane-based treatment (early-treatment)
| Arm | Measure | Value (NUMBER) |
|---|---|---|
| Early Treatment Change | Repeatability Coefficient (RC)Test-retest Metric for Reproducibility of ADC as Applied to Breast Tumors | 0.16 10E-3 mm/sec^2 |
Secondary
Within-subject Coefficient of Variation (wCV) Test-retest Metric for Reproducibility of ADC as Applied to Breast Tumors
within-subject standard deviation (wSD) Within-subject coefficient of variation (wCV): \[wCV = 100%\*wSD/mean\] Smaller values of wCV bounded for \[0,...) represent better agreement
Time frame: baseline (pre-treatment) or after 3 weeks of taxane-based treatment (early-treatment)
| Arm | Measure | Value (NUMBER) |
|---|---|---|
| Early Treatment Change | Within-subject Coefficient of Variation (wCV) Test-retest Metric for Reproducibility of ADC as Applied to Breast Tumors | 4.8 coefficient of variation |