Microinvasive breast cancer and the role of sentinel lymph node biopsy

Whether sentinel lymph node biopsy (SLNB) should be performed in patients with microinvasive breast cancer (MIBC) has been a matter of debate over the last decade. MIBC has a favorable prognosis and while metastasis to the axilla is rare, it can impact treatment recommendations. In this study we evaluated clinical and histological features in both MIBC and background DCIS including ER, PR, and HER-2, number of foci of MIBC, the extent of the DCIS, nuclear grade, presence of comedo necrosis, as well as surgical procedures, adjuvant treatment and follow up to identify variables which predict disease free survival (DFS), as well as the factors which influence clinical decision making. Our study included 72 MIBC patients with a mean patient follow-up time of 55 months. Three patients with MIBC had recurrence, and two deceased, leaving five patients in total with poor long-term outcomes and a DFS rate of 93.1%. Performing mastectomy, high nuclear grade, and negativity for ER and HER-2 were found to be associated with the use of SLNB, although none of these variables were found to be associated with DFS. One positive lymph node case was discovered following SLNB in our study. This suggests the use of SLNB may provide diagnostic information to some patients, although these are the anomalies. When comparing patients who had undergone SLNB to those which had not there was no difference in DFS. Certainly, the use of SLNB in MIBC is quite the conundrum. It is important to acknowledge that surgical complications have been reported, and traditional metrics used for risk assessment in invasive breast cancer may not hold true in the setting of microinvasion.

www.nature.com/scientificreports/ Therapeutic approaches can result in overtreatment of some patients with breast cancer. The Marmot Report published in 2012 acknowledge the negative effects of overtreatment to women's health 15 . There are many considerations for surgical interventions: poor cosmesis after surgery, chronic pain due to sentinel lymph node biopsy procedure in the axilla, and the possibility of no long-term outcome difference following procedure 13 . ER, PR, and HER-2 status have been extensively studied in invasive breast cancer, while less data is available regarding ER, PR and HER-2 in MIBC 6,16,17 . Although highly prevalent, there is little direct evidence that hormonal status is to be associated with improved long term outcomes in MIBC 18 . It is generally accepted that hormone receptor (HR) positive patients receive benefit from adjuvant endocrine therapy, however adjuvant chemotherapy in MIBC has been found to only improve the outcomes of ER(−)/PR(−) patients which did not overexpress Ki-67 19 .
Performing a risk assessment is important for guiding treatments and for MIBC there is limited information. Most research has not analyzed all pathology parameters together with clinical follow up. The principle aim of this study is to provide a comprehensive evaluation of MIBC and associated DCIS in relation to salient clinical and pathological variables including ER, PR, and HER-2 status, number of foci of MIBC, the size extent of the background DCIS and nuclear grade, presence of comedo-necrosis, as well as surgical procedures, adjuvant treatment and clinical follow-up to identify variables which predict disease free survival and influence clinical decision making.

Materials and methods
Case selection. Upon Lifespan Health System Institutional review board (IRB) approval (Lifespan IRB: 751551-10), a retrospective natural language search of the pathology database (Cerner CoPath) for patients over the age 18 years was performed from July 2010 to May 2020. Cases with a diagnosis of "microinvasive breast cancer" were retrieved, while definitively invasive carcinomas (> 1 mm) were excluded. Mammography was the single most common approach to breast cancer detection in this cohort of patients and all current AJCC protocols were followed according to the most current recommendations for the diagnosis of microinvasion (< 1 mm). Follow up data for death and recurrence was provided by the cancer registry at the Lifespan Health System. All methods were carried out in accordance with relevant guidelines and regulations and informed consent was not required by the Lifespan Health System Institutional review board secondary to the retrospective nature of this study.
Pathology examination. All cases were reviewed by two breast pathologists. Each focus of microinvasion was measured individually, with multiple foci not being added together. Figure 1 demonstrates histological findings in MIBC, including findings seen during immunostaining. The extent of DCIS was estimated by sequential sectioning of the slices or calculated from the ratio of number of blocks with DCIS to total slides or by estimating the number of centimeters of DCIS disease present comprising the number of blocks of DCIS multiplied by 4 mm in non-sequential setting 20 . We chose a 2 mm cutoff for close margins as used by other groups 21 . Immunohistochemistry. Anti-estrogen receptor (ER; 1:50; Dako, Santa Clara, CA; clone 1D5), progesterone receptor (PR; 1:400; Dako; clone 1A6), HER2/neu (Dako HercepTest), anti-p63 (1:100; Biocare; clone 4A4), anti-calponin (1:500; Dako; clone CALP), and anti-smooth muscle heavy chain (1:100, Cell Marque; clone CMC569) were used for immunohistochemistry. Immunoreactivity was detected using the Dako EnVision method according to the manufacturer's recommended protocol. Immunohistochemistry for ER and HER-2 was scored according to expression guidelines published by updated College of American Pathologists (CAP) and the American Society of Clinical Oncology (ASCO) 22 , which was recently updated in 2018 23 . ER and PR were reported as positive when greater than 1% of tumor nuclei showed staining and negative when less than 1% of tumor nuclei showed staining, HER-2 was reported as negative if scored as 1 + and positive when scored as 3 + . Tumors scored as 2 + underwent confirmatory testing with chromogenic in situ hybridization (CISH).
Chromogenic in situ hybridization. HER-2 CISH was performed by a VENTANA 4B5 Inform HER-2 dual-color on the BenchMark Ultra system (INFORM HER2 DNA dualcolor assay-Roche Tissue Diagnostics, VENTANA Medical Systems, SA) per the manufacturer-dictated protocol. Invasive breast cancer and synchronous DCIS components were scored and recorded separately according to the updated ASCO/CAP guidelines 23 : a HER-2 copy number of 6.0 or higher per cell or a HER-2:CEP17 ratio of 2 or higher was considered to represent HER-2 amplification. HER-2 copy numbers of < 4 signals per cell and HER-2/CEP17 ratios of < 2 was nonamplified.

Statistical analysis. The Fisher's exact test was used to determine differences in proportions and t-tests
were used to compare differences in means for parameters. All tests were 2-sided. Kaplan-Meier survival analysis was used to evaluate disease free survival (DFS) rate as a function of time, while the log-rank method was used to compare differences between groups. Univariate and multivariate analyses of DFS were performed by Cox proportional-hazard regression. For DFS variables were dichotomized as follows: patient age (≥ 50 vs < 50), SLNB (Performed vs Not Performed), surgery type (breast conserving therapy (BCT) vs mastectomy (MST), radiation status (+ vs −), DCIS size (≥ 25 mm vs < 25 mm), MIBC foci (≥ 2 vs 1), Margin status (+ /close vs −), DCIS nuclear grade (3 vs 1/2), pathologic necrosis (+ vs −), ER status (+ vs −), PR status (+ vs −) and HER-2 status (+ vs −). Fishers exact test and t-tests analyses were performed using SPSS statistics version 23.0.0.3. Cox-regression was performed for disease-free survival (DFS) on RStudio 2021.09.1 + 372 "Ghost Orchid". A p-value < 0.05 was considered statistically significant.
The mean DCIS size was 31 mm and nuclear grading was as followed: grade 1 (7), grade 2 (21) Clinicopathological features. Many features were found to be associated with treatment/management decisions and long-term outcomes. First, SLNB was more common in patients in younger patients (P = 0.001) and in those who underwent mastectomy (P = 0.004), and RT (P = 0.025). SLNB was also more commonly Table 1. Patient characteristics and pathology findings based on treatment in microinvasive breast cancer. SLNB sentinel lymph node biopsy, BCT breast conservative therapy, MST mastectomy, RT radiation therapy, SEM standard error of the mean, SD standard deviation. The bold indicated statistical significance. www.nature.com/scientificreports/ used in larger DCIS size (P = 0.012), DCIS with high nuclear grade (P = 0.023), and in patients who were ER (P = 0.044) and PR (P = 0.048) positive. The use of BCT was more commonly followed with adjuvant RT when compared to patients which had mastectomy (P = 0.001) and with a larger DCIS size (P = 0.001). RT was found to be associated with less extensive DCIS (P = 0.014). The remaining clinicopathological features were not statistically significant.
Disease free survival. DFS was defined as by the presence of recurrence or death. The average mean patient follow-up time was 55 months. Follow up for those without SLNB was 47 month, and 61 months for patients who underwent SLNB. Three patients with MIBC had recurrence and two deceased related to breast cancer mortality, leaving 5 patients in total with poor DFS and a DFS rate of 91.7%. The characteristics for these patients are presented in Table 2.
None of the following variables were significant following univariate or multivariate cox regression for DFS (P > 0.05). Results for Cox-regression can be found in Supplemental Table S1, while Fisher's exact and t-test results can be found in Supplemental Table S2.

Discussion
The present work evaluated clinical parameters including SLNB, surgical type, radiation status, and long-term outcomes in relation to histological features in MIBC and associated DCIS. Overall, MIBC was found to have a favorable prognosis and a DFS rate of 93.1%, congruent with other studies which have shown recurrence free survival rates between 90 and 97% 6,9,10,24 . These studies had an average follow up time of 64.9 months, an overview of all MIBC publications based on the 1 mm AJCC cutoff is summarized in Table 3.
In the present study, the use of SLNB was more often confined to patients undergoing mastectomy and was also performed more commonly in the background of DCIS showing high nuclear grades with ER negative and HER-2 positive receptor status. Only 1 of 44 patients (2.3%) had positive lymph nodal metastasis (> 0.2 mm) following SLNB and no variable was able to statistically predict nodal positivity. A finding possibly secondary to the low propensity of lymph node metastasis in MIBC. Importantly, when comparing patients who had undergone SLNB to those that had not, there was no difference in long term outcome in our study. Although 4 out of 5 patients (80%) who experienced poor DFS outcomes did not receive radiotherapy.
Despite the lack of differences in long term outcome following SLNB, surgical complications associated with SLNB biopsy have been reported. In a large study of 5327 patients performed by Wilke et al. 25 , complications included axillary wound infection (1.0%), axillary seroma (7.1%), and axillary hematoma (1.4%). For older patients, SLNB was also found to be associated with an increased incidence of axillary seroma.
We also observed HER-2 positivity to be relatively common in MIBC, and when compared with pure DCIS, numerous studies have reported a higher rate of HER-2 overexpression in MIBC than in both invasive carcinomas and DCIS 6,12 . This is counter intuitive, as HER-2 amplification has been traditionally seen to occur more often in DCIS than in invasive carcinoma 26 . Zhang et al. 16 demonstrated HER-2 positivity to be associated with high-grade morphologic features, but not nodal metastasis or worse outcomes. We also did not find HER-2 overexpression to be associated with recurrence in MIBC, despite the fact that HER-2 has been demonstrated to be an independent high risk predictor of early recurrence in invasive breast cancer 27 . These findings could be secondary to good overall outcomes in MIBC, as well as the small patient population in our study. Nonetheless, the role of HER-2 targeted therapy in MIBC will need to be validated in the clinical setting by future studies.
In the present study we analyzed clinicopathological parameters such as the number of microinvasive foci, and both the extent and histology of background DCIS, however none of these were found to be associated with long term clinical outcomes. Although some studies have found patients with multiple foci of microinvasion to have worse DFS outcomes 28 , a large study of 414 patients performed by Matsen et al. 29 showed that there was no higher risk of nodal involvement for patients with ≥ 2 foci of microinvasion when compared to 1 focus.
In a recent study from the Surveillance Epidemiology and End Results (SEER) database between 2003 and 2015, Chen et al. 30 examined nodal metastasis, axillary surgery, and prognosis in 11,692 MIBC patients. Multivariate analyses showed that nodal metastasis was the best survival predictor; however, axillary lymph node dissection did not demonstrate a statistically significant survival benefit at 10-years. www.nature.com/scientificreports/ There were several pitfalls in the present study. This was a retrospective study which always carries the risk of selection bias and the follow up time was relatively short. We also did not perform molecular testing and due to the low incidence of MIBC we were unable to increase our cohort size. Multi-institution collaboration will be important for future studies.
In summary, when comparing patients who had undergone SLNB to those who had not we found no difference in long term clinical outcomes. The decision to undergo SLNB in MIBC should be made with the knowledge that surgical complications are reported, and traditional metrics for risk stratification and treatment decision making in invasive mammary carcinoma may not hold true in the setting of microinvasion.

Data availability
The datasets used and analyzed during the current study are available from the corresponding author upon reasonable request.