Breast Cancer With Extensive Peritumoral Vascular Invasion
A total of 2,116 consecutive patients affected by IBC were collected between January 1992 and December 2006 and included in the database of the Department of Pathology of the G.B. Rossi Hospital in Verona, Italy. Data on the patient's medical history, surgery, pathologic evaluation, and results of staging procedures (bone scan, chest film, and upper abdominal ultrasound examination) were evaluated. Specimens removed through surgical biopsy or mastectomy were retrospectively reviewed by breast pathologists (E.M. and A.R.). Pathologic assessment included the primary tumor size, histologic type, histologic grade, and axillary lymph nodes status.
Clinical and/or histopathologic criteria were adopted to include a patient in the IBC group. A histopathologic diagnosis of IBC was made when neoplastic emboli were observed within dermal lymphatic spaces. Clinical evidence of IBC consisted of diffuse erythema, peau d'orange, edema, warmth, tenderness, breast enlargement, and diffuse induration of the breast on palpation, as described by Haagensen.
In the current study, only strict morphologic criteria were adopted to define the phenomenon of lymphovascular invasion (LVI) by tumor cells. On morphology, endovascular neoplastic emboli were considered "true" only when neoplastic cells were in channel spaces lined by endothelial cells Image 1; all other spaces without certain peripheral endothelium were considered LVI mimickers Image 2. No qualitative distinction, "lymphatic," "venous," or "arterial" invasion, or immunohistochemical staining were applied. Immunohistochemical staining for lymphatic and venous channels using D2–40 and CD31, respectively, have been proposed to solve the problem concerning the retraction artifacts and for determining the type of lining endothelium. Although immunostains may be helpful in confirming LVI, problems related to cross-reactivity and false-negative immunostaining for lymphatic endothelia have not been completely overcome, and the adoption of strict morphologic criteria with conventional H&E stain for LVI remain of clinical value in identifying high-risk patients.
(Enlarge Image)
Image 1.
Lymphovascular channel with the lumen partially occupied by an aggregate of tumor cells; note the inner channel wall lined by flattened endothelial cells (H&E, ×60).
(Enlarge Image)
Image 2.
Lymphovascular invasion mimicker. Artifactual retraction of stroma around tumor cells. The contour of tumor cell aggregate conforms exactly to the shape of the space in which it lies (H&E, ×40).
The presence of neoplastic emboli in the lumen of peritumoral lymphovascular spaces, evaluable in more than one tumor block on H&E-stained slides, poised at more than one high power microscopic field away from the boundaries of the main tumor and visible in more than 10 vascular spaces was consistent with a diagnosis of IBC with ePVI (ePVI-BC).
A casual subset of patients affected by primary IBC without histologic criteria of IBC and ePVI-BC represented the control group-BC. Patients included in the control group were casually recruited among the first 10 consecutive breast cancer cases diagnosed each year between 1992 and 2006.
Estrogen receptor (ER) and progesterone receptor (PR) status; Ki-67 labeling index (Ki-67), as assessed with MIB1 monoclonal antibodies; and Her2/neu overexpression (HER2) were evaluated immunohistochemically on formalin-fixed, paraffin-embedded tissue blocks. ER and PR status was recorded as positive when nuclear staining was identified in more than 5% of neoplastic cells. Ki-67 labeling index was stratified as low when the Ki-67 staining score was 1% to 10% of neoplastic cells, medium with 11% to 25%, and high with more than 25% of neoplastic cells. HER2 was graded from no staining (0) to weak (1+), moderate (2+), and intense membrane staining (3+) in more than 10% of tumor cells in the tissue section, according to HER2 testing guidelines of the American Society of Clinical Oncology and College of American Pathologists. A score of 2+ was used as the cutoff point for fluorescence in situ analysis.
The clinical files were examined for treatment received by patients as systemic adjuvant therapy and surgical treatment.
The Fisher exact and χ tests were used to assess the association between categorical and ordinal variables in the three different groups of invasive cancer (IBC, ePVI-BC, and control group-BC). Disease-free survival (DFS) was defined as the length of time from the date of surgery and any relapse, the appearance of a second primary cancer, death, or the date of last follow-up visit. Overall survival (OS) was determined as the time of surgery to the date of death from any cause, or the date of the last follow-up visit. Survival plots according to age were drawn using the Kaplan-Meier method. All analyses were performed with STATA 13 (StataCorp, College Station, TX). A P value less than .05 was considered statistically significant.
Materials and Methods
Patients
A total of 2,116 consecutive patients affected by IBC were collected between January 1992 and December 2006 and included in the database of the Department of Pathology of the G.B. Rossi Hospital in Verona, Italy. Data on the patient's medical history, surgery, pathologic evaluation, and results of staging procedures (bone scan, chest film, and upper abdominal ultrasound examination) were evaluated. Specimens removed through surgical biopsy or mastectomy were retrospectively reviewed by breast pathologists (E.M. and A.R.). Pathologic assessment included the primary tumor size, histologic type, histologic grade, and axillary lymph nodes status.
Clinical and/or histopathologic criteria were adopted to include a patient in the IBC group. A histopathologic diagnosis of IBC was made when neoplastic emboli were observed within dermal lymphatic spaces. Clinical evidence of IBC consisted of diffuse erythema, peau d'orange, edema, warmth, tenderness, breast enlargement, and diffuse induration of the breast on palpation, as described by Haagensen.
In the current study, only strict morphologic criteria were adopted to define the phenomenon of lymphovascular invasion (LVI) by tumor cells. On morphology, endovascular neoplastic emboli were considered "true" only when neoplastic cells were in channel spaces lined by endothelial cells Image 1; all other spaces without certain peripheral endothelium were considered LVI mimickers Image 2. No qualitative distinction, "lymphatic," "venous," or "arterial" invasion, or immunohistochemical staining were applied. Immunohistochemical staining for lymphatic and venous channels using D2–40 and CD31, respectively, have been proposed to solve the problem concerning the retraction artifacts and for determining the type of lining endothelium. Although immunostains may be helpful in confirming LVI, problems related to cross-reactivity and false-negative immunostaining for lymphatic endothelia have not been completely overcome, and the adoption of strict morphologic criteria with conventional H&E stain for LVI remain of clinical value in identifying high-risk patients.
(Enlarge Image)
Image 1.
Lymphovascular channel with the lumen partially occupied by an aggregate of tumor cells; note the inner channel wall lined by flattened endothelial cells (H&E, ×60).
(Enlarge Image)
Image 2.
Lymphovascular invasion mimicker. Artifactual retraction of stroma around tumor cells. The contour of tumor cell aggregate conforms exactly to the shape of the space in which it lies (H&E, ×40).
The presence of neoplastic emboli in the lumen of peritumoral lymphovascular spaces, evaluable in more than one tumor block on H&E-stained slides, poised at more than one high power microscopic field away from the boundaries of the main tumor and visible in more than 10 vascular spaces was consistent with a diagnosis of IBC with ePVI (ePVI-BC).
A casual subset of patients affected by primary IBC without histologic criteria of IBC and ePVI-BC represented the control group-BC. Patients included in the control group were casually recruited among the first 10 consecutive breast cancer cases diagnosed each year between 1992 and 2006.
Estrogen receptor (ER) and progesterone receptor (PR) status; Ki-67 labeling index (Ki-67), as assessed with MIB1 monoclonal antibodies; and Her2/neu overexpression (HER2) were evaluated immunohistochemically on formalin-fixed, paraffin-embedded tissue blocks. ER and PR status was recorded as positive when nuclear staining was identified in more than 5% of neoplastic cells. Ki-67 labeling index was stratified as low when the Ki-67 staining score was 1% to 10% of neoplastic cells, medium with 11% to 25%, and high with more than 25% of neoplastic cells. HER2 was graded from no staining (0) to weak (1+), moderate (2+), and intense membrane staining (3+) in more than 10% of tumor cells in the tissue section, according to HER2 testing guidelines of the American Society of Clinical Oncology and College of American Pathologists. A score of 2+ was used as the cutoff point for fluorescence in situ analysis.
The clinical files were examined for treatment received by patients as systemic adjuvant therapy and surgical treatment.
Statistical Analysis
The Fisher exact and χ tests were used to assess the association between categorical and ordinal variables in the three different groups of invasive cancer (IBC, ePVI-BC, and control group-BC). Disease-free survival (DFS) was defined as the length of time from the date of surgery and any relapse, the appearance of a second primary cancer, death, or the date of last follow-up visit. Overall survival (OS) was determined as the time of surgery to the date of death from any cause, or the date of the last follow-up visit. Survival plots according to age were drawn using the Kaplan-Meier method. All analyses were performed with STATA 13 (StataCorp, College Station, TX). A P value less than .05 was considered statistically significant.
SHARE
