JOURNAL TRANSCRIPT
Histopathology 2015, 66, 577–586. DOI: 10.1111/his.12550
Ductal carcinoma in situ with distorting sclerosis on core biopsy may be predictive of upstaging on excision Laura L Walters, Judy C Pang, Lili Zhao1 & Julie M Jorns Department of Pathology, and 1Department of Biostatistics, University of Michigan, Ann Arbor, MI, USA Date of submission 27 June 2014 Accepted for publication 15 September 2014 Published online Article Accepted 17 September 2014
Walters L L, Pang J C, Zhao L & Jorns J M (2015) Histopathology 66, 577–586. DOI: 10.1111/his.12550
Ductal carcinoma in situ with distorting sclerosis on core biopsy may be predictive of upstaging on excision Aims: The aim of this study was to examine clinicopathological features of patients with core biopsy diagnoses of ductal carcinoma in situ (DCIS) that may predict invasion on subsequent excision, as upstaging has significant implications regarding the need for axillary staging via sentinel lymph node biopsy (SLNB). Methods and results: We identified 186 patients with a diagnosis of DCIS as the highest-stage lesion on core biopsy. Pathological and clinical features were assessed via slide and chart review, respectively. Distorting sclerosis was defined as irregular angulation of glands involved by DCIS but lacking definite invasion according to histology and/or immunohistochemical staining for myoepithelial markers. Thirty-two of 186 (17.2%) cases had upstaging to
either microinvasive (nine) or invasive (23) ductal carcinoma. SLNB was performed in 29 of 32 (90.6%) cases with upstaging and in 55 of 154 (35.7%) cases without (P < 0.0001). Upstaging was significantly associated with the presurgical variables of radiological mass (P = 0.009) and distorting sclerosis (P = 0.0005) and the postsurgical feature of multifocality (P < 0.0001). Conclusions: Sentinel lymph node biopsy is frequently performed for patients with upstaging from DCIS on core biopsy to microinvasive or invasive carcinoma on excision. DCIS with distorting sclerosis without definite invasion on core biopsy may be predictive of upstaging. This feature may be useful in selecting patients to undergo SLNB at the time of excision to avoid reoperation.
Keywords: breast, core biopsy, distorting sclerosis, ductal carcinoma in situ
Introduction Percutaneous core biopsies of suspicious breast lesions have replaced excisional biopsies in many cases, because they are generally safer, more cost-effective, and less emotionally and physically disfiguring.1 However, despite the advantages of minimally invasive biopsies, core biopsy is inherently limited in the amount of tissue obtained for pathological evaluation, resulting in the possibility of missing invasive disease and sampling only in-situ disease. On average, 6.6% Address for correspondence: Dr J M Jorns, Department of Pathology, University of Michigan, 1500 E. Medical Center Dr. 2G332 UH, Ann Arbor, 48109-0054 MI, USA. e-mail:
[email protected]. edu © 2014 John Wiley & Sons Ltd.
and 22% of women diagnosed with ductal carcinoma in situ (DCIS) on core biopsy are upstaged to microinvasive and invasive ductal carcinoma, respectively, on subsequent excision, with an overall range of upstaging of 8–47%.2–18 Several studies have investigated clinicopathological features in patients with core biopsy diagnoses of DCIS, in an attempt to predict invasion on subsequent excision, as this has significant implications regarding the need for axillary staging via sentinel lymph node biopsy (SLNB). Currently, SLNB is the standard practice in patients with invasive disease and clinically negative axillary lymph nodes. However, the use of SLNB in patients with DCIS on core biopsy is controversial. The American Society of Clinical Oncology recently published clinical practice guidelines on the use of SLNB for patients
578
L L Walters et al.
with early-stage breast cancer.19 With regard to DCIS diagnosed by core biopsy, the panel recommends that women should be offered SLNB in the following circumstances: when a mastectomy will be the mode of definitive excision, as SLNB will not be technically feasible afterwards; when a mass highly suspicious for invasive carcinoma is either palpable or noted radiologically; or when the area of DCIS is ≥50 mm.19 Because reoperation causes additional physical and emotional stress for the patient, is technically more challenging, and imposes additional financial burdens on the patient and healthcare system, SLNB should be performed at the time of excision if indicated. However, SLNB should not be performed in all cases of pure DCIS, owing to the low, but real, probability of complications, including allergic reaction to the dye used, lymphoedema, infection, seroma, longer duration of anaesthesia, and injury to regional nerves, resulting in sensory and motor deficits.19 Ideally, patients showing features suggestive of invasive disease should be selected to undergo SLNB at the time of excision. To date, few features have consistently surfaced as strong predictors of upstaging on excision. The clinicopathological features that are most commonly associated with upstaging to invasive disease include the presence of a palpable or radiological mass, and core biopsy pathology of high nuclear grade DCIS with comedonecrosis.3,5,6,13,15,17,18 In this study, we evaluated a combination of clinical and pathological features in core biopsies with DCIS in an attempt to determine novel features that may predict upstaging to invasive carcinoma on excision.
Materials and methods After Institutional Review Board approval, a retrospective search (2006–2011) of the University of Michigan pathology electronic database identified 188 female patients with a diagnosis of DCIS as the highest-stage lesion on breast core biopsy. Only patients who underwent subsequent definitive excision (breast conservation therapy or mastectomy) at the University of Michigan were included. Core biopsies were either ultrasound-guided or stereotacticguided vacuum-assisted core needle biopsies. Clinical characteristics including patient age, the presence of a radiological mass and/or radiological calcifications,
biopsy method, number of sites biopsied, number of biopsy samples per site and timing of SLNB were obtained from patient electronic medical records. Diagnosis, size, multifocality (defined as two or more discrete foci at least 20 mm apart), the presence of calcifications in DCIS and the pathological status of sentinel lymph nodes [negative, micrometastasis (≤2 mm) and macrometastasis (>2 mm)] were obtained from pathology reports. All original haematoxylin and eosin-stained slides and available immunohistochemical (IHC) stains performed at the time of diagnosis were re-reviewed by two pathologists (J.P. and J.J.) with experience in breast pathology. Two cases with microinvasion confirmed by myoepithelial IHC staining were excluded. Histopathological features, including linear involvement of core biopsy, highest nuclear grade, presence of central necrosis, primary architectural pattern, cancerization of lobules, degree of periductal chronic inflammation, expression of oestrogen receptor (ER) and progesterone receptor (PR), presence of periductal fibrosis, and presence of distorting sclerosis, were included in the analysis. Focal was defined as involvement of one to three ducts, and extensive as involvement of more than three ducts. Distorting periductal sclerosis was defined as periglandular desmoplasia or fibrosis resulting in irregular angulation of glands involved by DCIS (Figure 1A–C), whereas periductal fibrosis was defined as a concentric rim of fibrosis encircling the duct without distortion of the lumen. Distorting sclerosis was defined as ‘severe’ when the distortion of the glands was marked and concerning for or difficult to discern from microinvasion via histological examination alone. Nuclear grade was determined according to consensus committee guidelines.20 ER and PR immunostains were reviewed and categorized as positive (>10% staining), weak (1–10% staining), or negative (