Reliability of Needle Core Biopsies of Large Breast Tumors
Reliability of Needle Core Biopsies of Large Breast Tumors
All patients 18 years or older who had been treated for newly diagnosed EBC in Henri Becquerel Center, Rouen, France, between January 2008 and December 2011 were retrospectively screened. Invasive breast carcinomas 2 cm or larger (pT2-pT4c) with available pretherapeutic NCB specimens and WSS and that were treated first by surgery (mastectomy or lumpectomy) without neoadjuvant treatments were eligible. Inflammatory, multifocal, metastatic, and in situ or microinvasive breast cancers were excluded. For each patient, informed consent for biomedical studies on histologic samples was collected at the time of surgery.
Sampling. NCBs were performed by a surgeon or a radiologist at the time of initial diagnosis under local anesthesia mostly with a 14-gauge needle. All samples were immediately fixed (for 24 hours) and shipped with minimal delay to the Department of Pathology. WSS (lumpectomy or mastectomy) were incised fresh and formalin fixed after gross examination within a maximum of 30 minutes, for at least 48 hours, depending on the size. The tumor size was assessed on the fresh surgical specimens.
Tissue Processing. Formalin-fixed, paraffin-embedded (FFPE) samples relative to the NCB specimens and WSS were collected, and data regarding preanalytic procedures (surgeon or radiologist, delay of shipment to the laboratory, fixation time) were registered. H&E staining and IHC for the ER, PR, and HER2 were performed using FFPE slides for NCB specimens and WSS, in contrast to analysis of Ki-67, cyclin A, cytokeratin 5 (CK5), and epidermal growth factor receptor (EGFR), which were performed using slides for NCB specimens and a tissue microarray (TMA) for the WSS. All antibody references are detailed in Supplemental Table 1 (supplemental material can be found at http://bit.ly/PetrauOct15). The TMA was constructed with six spots (0.6-mm–thick sections) focusing on the invasive component for each tumor, plus two normal breast spots for external controls. If fewer than two spots per tumor were interpretable (with at least 200 tumor cells/spot), or in cases of discrepancies with the NCB specimens, additional IHC analysis was performed on individual WSS sections.
Histologic Analysis and Definition of Intratumor Heterogeneity. All samples were independently analyzed twice by two well-trained pathologists (A.B. and J.-M.P.), and if discordant, they were submitted to a third pathologist (L.V.).
Data (number and length of biopsy specimens, corresponding length of tumor tissue), the presence of biopsy artifacts (crushing or fixation, defined by cramping interpretation), histologic type, and SBR grade (including MAI, as defined by the Nottingham Grading System) were assessed on H&E-stained sections for NCB specimens and WSS. Intratumor heterogeneity was assessed only for WSS and was defined as the presence of mixed histologic types or tumor grade within the same sample.
IHC Interpretation. IHC evaluation was restricted to nuclear staining for ER, PR, and Ki-67; cytoplasmic staining for CK5; and membranous staining for EGFR and HER2. HER2 assessment was performed according to the 2007 American Society of Clinical Oncology/College of American Pathologists (ASCO/CAP) recommendations. All HER2 2+ samples by IHC were tested for amplification by chromogenic in situ hybridization (CISH) (DuoCish SK108 kit; DAKO, Glostrup, Denmark). The positivity cutoffs for each factor are detailed in Table 1. Intrinsic subtypes were defined according to the 2011 St Gallen guidelinesTable 2. The EGFR and CK5 assessments were used to define the basal-like phenotypes among the triple-negative tumors.
Definitions of Concordance Between NCB Specimens and WSS. Histologic type was considered concordant if the same type was determined by NCB specimens and the corresponding WSS. SBR grade was concordant if the same grade (I, II, or III) was defined by NCB specimens and WSS. ER, PR, HER2, Ki-67, cyclin A, EGFR, and CK5 were concordant if NCB specimens and WSS resulted in the same status (positive/negative).
Definition of Major Discordances Between NCB Specimens and WSS. In cases of discrepancies between the NCB specimens and WSS, a major discordance was retained when it potentially affected the treatment according to French recommendations (indication of adjuvant/neoadjuvant chemotherapy, hormonotherapy, or trastuzumab). Consequently, major discordances could be retained in cases of up- or downgrading from SBR I to III or conversion of the HR or HER2 status between the NCB specimens and WSS.
Putative Factors for Discordance. We hypothesized that the tumor size (20–29 mm vs 30–39 mm vs ≥40 mm), number of biopsies (1 or 2 vs 3 or 4 vs 5 or more), ratio number of biopsies/tumor size, presence of intratumor heterogeneity by WSS, or presence of biopsy artifacts constituted putative factors for discrepancies.
Definition of the Interobserver Concordance. An interobserver concordance was defined as the same determination for each histologic or IHC parameter, by NCB specimens or WSS, by the two readers.
Statistical Analysis. Analysis of concordance was performed using the Cohen's κ test with the MedCalc computer software (MedCalc Software, Ostend, Belgium). A concordance was defined as excellent if κ > 0.9, strong if 0.8 ≤ κ ≤ 0.9, moderate if 0.6 ≤ κ < 0.8, weak if 0.4 ≤ κ < 0.6, and minimal if less than 0.4.
Materials and Methods
Patients
All patients 18 years or older who had been treated for newly diagnosed EBC in Henri Becquerel Center, Rouen, France, between January 2008 and December 2011 were retrospectively screened. Invasive breast carcinomas 2 cm or larger (pT2-pT4c) with available pretherapeutic NCB specimens and WSS and that were treated first by surgery (mastectomy or lumpectomy) without neoadjuvant treatments were eligible. Inflammatory, multifocal, metastatic, and in situ or microinvasive breast cancers were excluded. For each patient, informed consent for biomedical studies on histologic samples was collected at the time of surgery.
Methods
Sampling. NCBs were performed by a surgeon or a radiologist at the time of initial diagnosis under local anesthesia mostly with a 14-gauge needle. All samples were immediately fixed (for 24 hours) and shipped with minimal delay to the Department of Pathology. WSS (lumpectomy or mastectomy) were incised fresh and formalin fixed after gross examination within a maximum of 30 minutes, for at least 48 hours, depending on the size. The tumor size was assessed on the fresh surgical specimens.
Tissue Processing. Formalin-fixed, paraffin-embedded (FFPE) samples relative to the NCB specimens and WSS were collected, and data regarding preanalytic procedures (surgeon or radiologist, delay of shipment to the laboratory, fixation time) were registered. H&E staining and IHC for the ER, PR, and HER2 were performed using FFPE slides for NCB specimens and WSS, in contrast to analysis of Ki-67, cyclin A, cytokeratin 5 (CK5), and epidermal growth factor receptor (EGFR), which were performed using slides for NCB specimens and a tissue microarray (TMA) for the WSS. All antibody references are detailed in Supplemental Table 1 (supplemental material can be found at http://bit.ly/PetrauOct15). The TMA was constructed with six spots (0.6-mm–thick sections) focusing on the invasive component for each tumor, plus two normal breast spots for external controls. If fewer than two spots per tumor were interpretable (with at least 200 tumor cells/spot), or in cases of discrepancies with the NCB specimens, additional IHC analysis was performed on individual WSS sections.
Histologic Analysis and Definition of Intratumor Heterogeneity. All samples were independently analyzed twice by two well-trained pathologists (A.B. and J.-M.P.), and if discordant, they were submitted to a third pathologist (L.V.).
Data (number and length of biopsy specimens, corresponding length of tumor tissue), the presence of biopsy artifacts (crushing or fixation, defined by cramping interpretation), histologic type, and SBR grade (including MAI, as defined by the Nottingham Grading System) were assessed on H&E-stained sections for NCB specimens and WSS. Intratumor heterogeneity was assessed only for WSS and was defined as the presence of mixed histologic types or tumor grade within the same sample.
IHC Interpretation. IHC evaluation was restricted to nuclear staining for ER, PR, and Ki-67; cytoplasmic staining for CK5; and membranous staining for EGFR and HER2. HER2 assessment was performed according to the 2007 American Society of Clinical Oncology/College of American Pathologists (ASCO/CAP) recommendations. All HER2 2+ samples by IHC were tested for amplification by chromogenic in situ hybridization (CISH) (DuoCish SK108 kit; DAKO, Glostrup, Denmark). The positivity cutoffs for each factor are detailed in Table 1. Intrinsic subtypes were defined according to the 2011 St Gallen guidelinesTable 2. The EGFR and CK5 assessments were used to define the basal-like phenotypes among the triple-negative tumors.
Definitions of Concordance Between NCB Specimens and WSS. Histologic type was considered concordant if the same type was determined by NCB specimens and the corresponding WSS. SBR grade was concordant if the same grade (I, II, or III) was defined by NCB specimens and WSS. ER, PR, HER2, Ki-67, cyclin A, EGFR, and CK5 were concordant if NCB specimens and WSS resulted in the same status (positive/negative).
Definition of Major Discordances Between NCB Specimens and WSS. In cases of discrepancies between the NCB specimens and WSS, a major discordance was retained when it potentially affected the treatment according to French recommendations (indication of adjuvant/neoadjuvant chemotherapy, hormonotherapy, or trastuzumab). Consequently, major discordances could be retained in cases of up- or downgrading from SBR I to III or conversion of the HR or HER2 status between the NCB specimens and WSS.
Putative Factors for Discordance. We hypothesized that the tumor size (20–29 mm vs 30–39 mm vs ≥40 mm), number of biopsies (1 or 2 vs 3 or 4 vs 5 or more), ratio number of biopsies/tumor size, presence of intratumor heterogeneity by WSS, or presence of biopsy artifacts constituted putative factors for discrepancies.
Definition of the Interobserver Concordance. An interobserver concordance was defined as the same determination for each histologic or IHC parameter, by NCB specimens or WSS, by the two readers.
Statistical Analysis. Analysis of concordance was performed using the Cohen's κ test with the MedCalc computer software (MedCalc Software, Ostend, Belgium). A concordance was defined as excellent if κ > 0.9, strong if 0.8 ≤ κ ≤ 0.9, moderate if 0.6 ≤ κ < 0.8, weak if 0.4 ≤ κ < 0.6, and minimal if less than 0.4.
Source...