The Use of Sentinel Lymph Node Biopsy in Patients Undergoing Mastectomy for DCIS

Abstract

Background:

Current guidelines do not recommend routine sentinel node biopsy (SLNB) for ductal carcinoma in situ (DCIS), except in the setting of mastectomy or microinvasive disease. This study aimed to evaluate national SLNB utilization in women undergoing upfront mastectomy for DCIS, identify predictors of SLNB utilization, and determine the percentage with a positive SLNB.

Methods:

A retrospective cohort analysis was performed using the NCDB of women with clinical DCIS who underwent upfront mastectomy between 2012–2017. Demographic and clinicopathologic variables were compared between patients who underwent SLNB and those who did not. Multivariate logistic regression models were used to identify factors associated with SLNB utilization and positive SLNB.

Results:

38,973 patients met inclusion criteria: 34,231 (88%) underwent SLNB and 4,742 (12%) had no surgical axillary staging. Most patients were age 50–69 (51%), non-Hispanic White (71%), with private insurance (66%). On multivariate analysis, older patients were less likely to receive SLNB (p<0.01), while patients with higher grade DCIS were more likely to undergo SLNB (p<0.01). In those who underwent SLNB (n=34,231), only 1,149 (3.4%) had nodal involvement. Non-Hispanic Black patients had increased odds of a positive SLNB (p<0.01), while those with estrogen receptor positive disease were less likely to be node positive (OR 0.68, p<0.001).

Conclusions:

While 88% of patients had a SLNB, only 3.4% were found to be node positive. Given this low rate, it is reasonable to consider SLNB omission in select patients with low grade, hormone receptor positive DCIS undergoing upfront mastectomy.

Micro Abstract:

The NCDB was used to examine SLNB utilization in patients with DCIS undergoing upfront mastectomy. Findings indicated low rates of nodal involvement and that hormone receptor-negative and high-grade DCIS were associated with positive SLNB, suggesting it is reasonable to consider SLNB omission for select populations.

Introduction

Ductal carcinoma in situ (DCIS) accounts for approximately 1 in 5 new breast cancers with >50,000 cases diagnosed annually in the United States.¹ While DCIS is often described as “non-invasive” or “pre-invasive breast cancer,” the local treatment paradigm for DCIS mirrors treatment for invasive breast cancer.² First line treatment for DCIS includes breast conserving surgery or mastectomy, and axillary staging may be warranted in select circumstances.³ While the benefits of surgical axillary staging for invasive cancer is well established, its utility for patients with DCIS remains controversial.^4–7 The benefits of finding axillary lymph node metastasis in patients with clinical DCIS must be weighed against the risks of surgical axillary staging including breast cancer related lymphedema, pain, and paresthesia in the ipsilateral upper extremity.^8–9

The National Comprehensive Cancer Network (NCCN) provides detailed guidelines on indications for surgical axillary staging in patients with DCIS. Currently, NCCN guidelines do not recommended the routine use of sentinel lymph node biopsy (SLNB) in patients undergoing breast conserving therapy for DCIS given that if invasive cancer is found on final surgical pathology, SLNB can be performed as a secondary procedure. However, NCCN guidelines recommend SLNB should be considered in patients with DCIS undergoing mastectomy given that mastectomy can compromise the success of axillary mapping at a future operation due to lymphatic disruption.² Recent literature identified that surgical axillary staging practice patterns for DCIS patients are inconsistent with guideline recommendations.^6,10 Currently, neither the rate of SLNB utilization in patients with DCIS undergoing mastectomy nor which factors are associated with a higher likelihood of undergoing SLNB are known. The objectives of this study were to evaluate national rates of SLNB in patients undergoing upfront mastectomy for DCIS, identify which factors are predictive of SLNB utilization, and determine both the percentage who have a positive SLNB and factors predictive of a positive SLNB.

Materials and Methods

A retrospective cohort analysis was performed using the National Cancer Database (NCDB). The NCDB, a joint project between the American College of Surgeons and the American Cancer Society, contains hospital registry data from Commission on Cancer-accredited facilities in the United States. The database includes >70% of newly diagnosed cancer cases with information on patient demographics, cancer characteristics, treatment, and outcomes (NCDB). As NCDB data is de-identified, this study was reviewed and deemed exempt by our institution’s Institutional Review Board.

Female patients diagnosed with clinical DCIS from 2012–2017 who underwent upfront mastectomy were included. DCIS was defined as patients who were coded as having both clinical Tis and stage 0 breast cancer. All patients within the cohort had breast cancer as their only documented malignancy and received most, if not all, of their treatment at the reporting facility.

Patient demographics, clinicopathologic, and treatment variables were compared between patients who received a SLNB with upfront mastectomy and those who had no surgical axillary staging. Two-sample t-tests were used to examine continuous variables and chi-squared test for categorical variables. A multivariate logistic regression model was then used to examine factors predictive of SLNB utilization. A subgroup analysis was conducted to describe patients with a positive SLNB. Lastly, a multivariate logistic regression analysis was used to identify factors associated with positive SLNB. A likelihood ratio test (LRT) was used to evaluate an interaction between variables. STATA version 18 was used for the statistical analysis.

Results

A total of 38,973 patients met study inclusion criteria: 34,231 (88%) underwent SLNB and 4,742 (12%) had no surgical axillary staging (Table 1). Most patients were age 50–69 years (51%), non-Hispanic White (71%), healthy (defined by a Charlson-Deyo comorbidity score ≤1) (97%), with grade 2 or 3 (72%), estrogen receptor-positive (ER-positive) (76%), progesterone receptor-positive (PR-positive)(62%) DCIS.

Table 1.

Patient demographic and clinicopathologic variables for female patients diagnosed with clinical DCIS from 2012–2017 who underwent upfront mastectomy and separated by type of axillary surgical staging.

	Total (N=38,973)	No SNLB (n=4,742)	SLNB (n=34,231)	p-value
Age (years)				<0.001
<50	13,626 (35%)	1,436 (30%)	12,190 (36%)
50–69	20,029 (51%)	2,387 (50%)	17,642 (51%)
≥70	5,318 (14%)	919 (20%)	4,399 (13%)
Race / Ethnicity				<0.001
NH White	27,761 (71%)	3,448 (73%)	24,313 (71%)
NH Black	5,192 (13%)	552 (12%)	4,640 (14%)
Hispanic	2,451 (6%)	313 (6%)	2,138 (6%)
Other	2,657 (7%)	288 (6%)	2,369 (7%)
Unknown	912 (3%)	141 (3%)	771 (2%)
Tumor Grade				<0.001
1	3,768 (10%)	580 (12%)	3,188 (9%)
2	12,992 (33%)	1,560 (33%)	11,432 (34%)
3	15,172 (39%)	1,361 (29%)	13,811 (40%)
Unknown	7,041 (18%)	1,241 (26%)	5,800 (17%)
Regional Lymph Nodes Examined				<0.001
0	4,851 (12%)	4,357 (92%)	494 (2%)
1–5	30,350 (78%)	283 (6%)	30,067 (88%)
≥10	1,039 (3%)	8 (0%)	1,031 (3%)
Unknown	2,733 (7%)	94 (2%)	2,639 (7)
Positive Regional Lymph Nodes*				0.013
0	32,944 (85%)	372 (8%)	32,572 (95%)
1–3	1,019 (3%)	3 (0%)	1,016 (3%)
≥4	133 (0%)	0 (0%)	133 (0%)
Unknown	4,877 (12%)	4,367 (92%)	510 (2%)
ER Status				<0.001
Positive	29,595 (76%)	3,472 (73%)	26,123 (76%)
Negative	7,396 (19%)	614 (13%)	6,782 (20%)
Unknown	1,982 (5%)	656 (14%)	1,326 (4%)
PR Status				<0.001
Positive	24,169 (62%)	2,897 (61%)	21,272 (62%)
Negative	10,869 (28%)	956 (20%)	9,913 (29%)
Unknown	3,935 (10%)	889 (19%)	3,046 (9%)
AJCC Pathologic N				<0.001
pN0	35,079 (90%)	2,728 (58%)	32,351 (95%)
pN1	1,017 (3%)	4 (0%)	1,013 (3%)
pN2	97 (0%)	0 (0%)	97 (0%)
pN3	38 (0%)	(0%)	38 (0%)
Unknown	2,742 (7%)	2,010 (42%)	732 (2%)
AJCC Pathologic Stage Group				<0.001
0	29,828 (77%)	3,807 (80%)	26,021 (76%)
1	6,443 (17%)	158 (4%)	6,285 (18%)
2	889 (2%)	9 (0%)	880 (3%)
3	174 (0%)	2 (0%)	172 (0%)
Unknown	1,639 (4%)	766 (16%)	873 (3%)
Chemotherapy				<0.001
Yes	2,138 (5%)	30 (1%)	2,108 (6%)
No	36,468 (94%)	4,668 (98%)	31,800 (93%)
Unknown	367 (1%)	44 (1%)	323 (1%)
Endocrine Therapy				<0.001
Yes	9,901 (26%)	881 (18%)	9,020 (27%)
No	27,367 (70%)	3,640 (77%)	23,727 (69%)
Unknown	1,705 (4%)	221 (5%)	1,484 (4%)
Radiation Therapy				0.207
Yes	1,312 (3%)	145 (3%)	1,167 (3%)
No	37,496 (96%)	4,578 (97%)	32,918 (96%)
Unknown	165 (1%)	19 (0%)	146 (1%)
Charlson Deyo Comorbidity score				0.554
0	32,711(84%)	3,957 (83%)	28,754 (84%)
1	4,958 (13%)	618 (13%)	4,340 (13%)
2	942 (2%)	126 (3%)	816 (2%)
≥3	362 (1%)	41 (1%)	321 (1%)
Primary payor				<0.001
Private	25,872 (66%)	2,882 (61%)	22,990 (67%)
Medicaid	2,412 (6%)	264 (6%)	2,148 (6%)
Medicare	9,206 (24%)	1,401 (30%)	7,805 (23%)
Other government	542 (1%)	73 (1%)	469 (1%)
Not insured	624 (2%)	68 (1%)	556 (2%)
Unknown	317 (1%)	54 (1%)	263 (1%)
Region				<0.001
New England	1,781 (5%)	237 (5%)	1,544 (5%)
Middle Atlantic	5,035 (13%)	563 (12%)	4,472 (13%)
South Atlantic	8,307 (21%)	1,032 (22%)	7,275 (21%)
East North Central	5,961 (15%)	723 (15%)	5,238 (15%)
East South Central	2,901 (8%)	428 (9%)	2,473 (7%)
West North Central	2,871 (7%)	305 (6%)	2,566 (8%)
West South Central	3,298 (8%)	425 (9%)	2,873 (8%)
Mountain	1,775 (5%)	248 (5%)	1,527 (5%)
Pacific	4,314 (11%)	550 (12%)	3,764 (11%)
Unknown	2,730 (7%)	231 (5%)	2,499 (7%)
Distance to treatment facility (miles)				0.674
0 – 20	25,100 (64%)	3,059 (64%)	22,041 (64%)
21 – 40	4,853 (13%)	600 (13%)	4,253 (13%)
41 – 60	1,745 (5%)	223 (5%)	1,522 (5%)
≥60	2,098 (5%)	242 (5%)	1,856 (5%)
Unknown	5,177 (13%)	618 (13%)	4,559 (13%)
Treatment facility				<0.001
Community cancer program	2,485 (6%)	517 (11%)	1,968 (6%)
Comprehensive community cancer program	16,022 (41%)	2,205 (46%)	13,817 (40%)
Academic/research program	11,689 (30%)	1,131 (24%)	10,558 (31%)
Integrated network cancer program	6,047 (16%)	658 (14%)	5,389 (16%)
Unknown	2,730 (7%)	231 (5%)	2,499 (7%)

Key: NH non-Hispanic, ER estrogen receptor, PR progesterone receptor, Positive regional lymph nodes is a subgroup of those individuals with regional lymph nodes removed and examined, AJCC Pathological N and Pathological Stage Group is based on AJCC Cancer Staging Manual 6^th and 7^th edition.

Compared to patients that did not undergo SLNB, those that underwent SLNB were frequently younger (<70 years of age) (87% vs 80%, p<0.001) and had grade 3 tumors (40% vs 29%, p<0.001). The majority of patients who underwent SLNB were treated at a comprehensive community cancer program (40%, p<0.001) and had private insurance (67%, p<0.001). Among patients who underwent SLNB (n=34,231), 7,165 (21%) were upstaged to invasive cancer on surgical pathology.

On multivariate analysis, several factors were predictive of SLNB utilization. Patients age 50–69 years were less likely to receive SLNB compared to those <50 (OR 0.85, 95% CI 0.76–0.94, p<0.01), as were those age ≥70 (OR 0.60, 95% CI 0.51– 0.71, p<0.01). Women with grade 2 or 3 DCIS were more likely to undergo SLNB compared to patients with grade 1 disease (OR 1.20, 95% CI 1.06–1.36 and OR 1.56, 95% CI 1.37–1.78, p<0.01, respectively) (Table 2). Patients who had PR-positive DCIS were less likely to undergo SLNB compared to those with PR-negative disease (OR 0.83, 95% CI 0.72– 0.95, p<0.01). Women receiving treatment in the Middle Atlantic, West North Central, and West South-Central United States were more likely to undergo SLNB compared to patients treated in other regions of the country (p<0.05). Additionally, patients treated at a comprehensive community program, academic/research program, or integrated network program were more likely to undergo SLNB compared to community programs (p<0.01).

Table 2.

Multivariate logistic regression analysis of factors associated with undergoing sentinel lymph node biopsy (n=22,642) in females with DCIS undergoing upfront mastectomy in the NCDB from 2012–2017.

Variable	Multivariable
	OR	95% CI	p-value
Age (years)
<50 (Reference)
50–69	0.85	0.76 – 0.94	0.002
≥70	0.60	0.51 – 0.71	<0.001
Race & Ethnicity
NH White (Reference)	1
NH Black	1.17	1.02 – 1.34	0.027
Hispanic	0.89	0.74 – 1.06	0.191
Other	1.05	0.88 – 1.24	0.610
Tumor Grade
Grade 1 (Reference)	1
Grade 2	1.20	1.06 – 1.36	0.004
Grade 3	1.56	1.37 – 1.78	<0.001
ER Status
Negative (Reference)	1
Positive	0.86	0.73 – 1.01	0.069
PR Status
Negative (Reference)	1
Positive	0.83	0.72 – 0.95	0.006
Charlson Deyo Comorbidity score
0 (Reference)	1
1	1.04	0.92 – 1.18	0.554
2	1.01	0.79 – 1.30	0.927
≥3	1.25	0.80 – 1.94	0.325
Insurance status
No insurance (Reference)	1
Private	0.79	0.52 – 1.20	0.262
Medicaid	0.68	0.44 – 1.07	0.097
Medicare	0.63	0.41 – 0.97	0.035
Other government	0.64	0.37 – 1.10	0.107
Facility location
New England (Reference)	1
Middle Atlantic	1.28	1.02 – 1.62	0.036
South Atlantic	1.18	0.95 – 1.47	0.134
East North Central	1.12	0.90 – 1.41	0.302
East South Central	0.86	0.67 – 1.09	0.206
West North Central	1.37	1.06 – 1.76	0.015
West South Central	1.30	1.02 – 1.65	0.038
Mountain	1.02	0.78 – 1.34	0.857
Pacific	1.05	0.83 – 1.32	0.677
Facility type
Community Cancer Program (Reference)	1
Comprehensive Community Cancer Program	1.56	1.35 – 1.80	<0.001
Academic / Research Program	2.25	1.93 – 2.63	<0.001
Integrated Network Cancer Program	2.36	1.97 – 2.82	<0.001
Distance to treating facility (miles)
0–20 (Reference)	1
20 – 40	1.05	0.93 – 1.19	0.453
40 – 60	0.92	0.76 – 1.11	0.401
≥ 60 miles	0.97	0.80 – 1.16	0.714

Key: NH non-Hispanic, CI confidence-interval, ER estrogen receptor, PR progesterone receptor

In patients who underwent SLNB (n=34,231), 1,149 (3.4%) were found to have nodal metastasis (Table 3) with >88% only having 1–3 positive nodes. Among patients with any nodal involvement, approximately 90% were age <70 years. For women age >70 who underwent SLNB, only 2.7% had a positive SLNB. Most patients with a positive SLNB were non-Hispanic White (67%), with higher grade (78%) and hormone-receptor positive disease (ER-positive 75%, PR-positive 63%).

Table 3.

Patient demographic and clinicopathologic variables for female patients diagnosed with clinical DCIS from 2012–2017 who underwent upfront mastectomy with SLNB and were found to have a positive SLN(s)

Variable	Total (N=1,149) n (%)
Age (years)
<50	537 (47%)
50–69	495 (43%)
≥70	117 (10%)
Race / Ethnicity
NH White	775 (67%)
NH Black	214 (19%)
Hispanic	89 (8%)
Other	49 (4%)
Unknown	22 (2%)
Number of Positive Regional Lymph Nodes
1–3	1,016 (88%)
≥4	133 (12%)
Tumor Grade
1	114 (10%)
2	465 (40%)
3	431 (38%)
Unknown	139 (12%)
ER Status
Positive	862 (75%)
Negative	282 (25%)
Unknown	5 (0%)
PR Status
Positive	727 (63%)
Negative	410 (36%)
Unknown	12 (1%)
Chemotherapy
Yes	770 (67%)
No	361 (31%)
Unknown	18 (2%)
Endocrine Therapy
Yes	783 (68%)
No	342 (30%)
Unknown	24 (2%)
Radiation Therapy
Yes	478 (41%)
No	663 (58%)
Unknown	8 (1%)

Key: NH non-Hispanic, Positive regional lymph nodes is a subgroup of those individuals with regional lymph nodes removed and examined, ER estrogen receptor, PR progesterone receptor.

Several factors were predictive of a having a positive SLNB at the time of upfront mastectomy. While non-Hispanic Black patients had a higher likelihood of a positive SLNB (OR 1.34, 95% CI 1.13–1.59, p=0.001), those with ER-positive disease were less likely to be node positive (OR 0.68, 95% CI 0.54 – 0.85, p=0.001). (Table 4). Grade affected the risk of a positive SLNB, but it varied by age. In the multivariate logistic regression model, a statistically significant interaction between age and grade was observed (p<0.001, LRT). For younger patients, higher grade disease was associated with a higher probability of nodal positivity while for older patients a higher grade was associated with a lower probability of nodal positivity (p<0.001). (Figure 1).

Table 4.

Multivariate logistic regression analysis of factors associated with a having a positive SLN(s) (n=28,021)

Variable	Multivariable
	OR	95% CI	p-value
Age	0.99	0.97 – 1.01	0.175
Tumor Grade
Grade 1 (Reference)	1
Grade 2	1.43	0.52 – 3.96	0.489
Grade 3	3.77	1.36 – 10.47	0.011
Tumor Grade x Age
Grade I (Reference)
Grade II	0.99	0.98 – 1.01	0.558
Grade III	0.97	0.95 – 0.99	0.001
Race and Ethnicity
NH White (Reference)	1
NH Black	1.34	1.13 – 1.59	0.001
Hispanic	1.17	0.92 – 1.49	0.207
Other	0.56	0.40 – 0.77	<0.001
ER Status
Negative (Reference)	1
Positive	0.68	0.54 – 0.85	0.001
PR Status
Negative (Reference)	1
Positive	0.80	0.65 – 0.97	0.974

Key: NH non-Hispanic, ER estrogen receptor, PR progesterone receptor

Figure 1.

Predicted probabilities of having a positive SLN(s) based on patient age and disease grade.

Discussion

In this retrospective analysis, we found the majority of patients with DCIS undergoing mastectomy underwent SLNB. We identified that patients were more likely to undergo SLNB if they were younger and if they had higher grade or hormone-receptor negative disease. Patients treated in the Middle Atlantic, West North Central, and West South-Central United States, and at an academic institution had a higher likelihood of undergoing SLNB compared to their counterparts. While SLNB was performed in 88% of patients, only 3.4% were found to have nodal metastasis; most of which had a low nodal burden. Non-Hispanic Black patients and those with hormone-receptor negative disease were more likely to have a positive SLNB. Additionally, a clear interaction was observed based on disease grade and patient age that affected the risk of a positive SLNB. This study is the first to examine modern national rates of SLNB utilization and demonstrate the low rate of nodal involvement in patients undergoing upfront mastectomy for DCIS.

Advances in screening mammography and subsequent improvements in screening modalities for breast cancer detection has led to a considerable increase in the detection of DCIS in recent decades, with a substantial increase in DCIS incidence among younger and Black women.^1–11 Previous NCDB data found that rates of SLNB utilization in women undergoing mastectomy for DCIS increased from 24.3% in 1998 to 77.1% in 2011 while rates of ALND simultaneously decreased from 50.0% to 16.3%.⁶ This more recent series identified that from 2012–2017, 88% received SLNB, which suggests an increase in concordance with NCCN guidelines over the past decade.²

Our findings identified that younger patients with higher-grade disease were more likely to receive SLNB in addition to those treated at an academic institution. These data are consistent with a previously published study conducted prior to 2012 which identified that younger patients treated at an academic institution with larger volume DCIS were more likely to undergo SLNB.⁶ These results are not surprising given that breast cancer tends to be more aggressive in younger women, high grade DCIS is more often associated with an upgrade to invasive cancer on surgical pathology, and academic centers are more likely to follow guideline concordant breast cancer treatment recommendations.¹²

Only 3.4% of patients who underwent SLNB were found to have nodal involvement and is comparable to previous reports.^3,7,13 Additionally, the majority of patients with nodal involvement had higher grade disease and further support the findings of smaller cohort studies that identified disease size, multifocal disease, and higher-grade DCIS to be associated with pathological nodal involvement.^7,14

Several guidelines recommend SLNB in patients undergoing mastectomy for DCIS due to lymphatic disruption and low likelihood of axillary mapping after mastectomy.^2,15–16 In the event invasive cancer is found on surgical pathology after mastectomy, most patients will subsequently require axillary lymph node dissection for nodal staging due high rates of failed axillary mapping. Given the low rate of nodal involvement in our series, it may be reasonable to omit SLNB in select patients with DCIS who undergo mastectomy, particularly older patients with low grade, hormone receptor positive disease who would meet the Choosing Wisely guidelines if found to have invasive disease. The Society of Surgical Oncology’s Choosing Wisely guidelines recommend the omission of routinely performing SLNB in women age ≥70 with clinically node negative, early stage, hormone receptor positive, HER2 negative invasive breast cancer¹⁷ given comparable outcomes regardless of surgical nodal staging.^17–18 In the present study, among women age >70 who underwent SLNB, only 2.7% were found to have a positive node. Additionally, patients with ER-positive disease were less likely to have a positive SLNB. Interestingly, we found that for older women higher grade disease did not confer an increased risk of a positive SLNB compared to younger women. Together these findings support that SLNB should be omitted in older patients with low grade hormone receptor positive DCIS undergoing upfront mastectomy and should be further evaluated through prospective investigation.

The low rate of nodal involvement in this cohort suggest that we may be overtreating the axilla in patients undergoing mastectomy for DCIS. SLNB alone carries a 6–7% lifetime risk of breast cancer related lymphedema and a 3–9% risk of paresthesia and is often underestimated by surgeons when discussing risk of SLNB with patients.^7,19–20 Additional complications include upper extremity pain, limitations in range of motion, and adverse reactions to mapping agents.^8–9 These lifelong complications impose significant morbidity for patients who may otherwise not have yielded much benefit from this procedure. Furthermore, it is estimated that the cost of performing SLNB exceeds $6,000, leading to significant patient and healthcare expenditure.²¹ Judicial use of SLNB in this population has the potential to decrease post-operative morbidity and decrease unnecessary healthcare costs and utilization.

Given the low rates of upstage to invasive disease, the use of SLNB in all patients undergoing mastectomy for all DCIS may be unwarranted. However, it appears that SLNB continues to play an integral role for patients at a higher risk of upstaging on final surgical pathology. We found 21% of patients were upstaged to invasive cancer on final surgical pathology, particularly in those with higher grade disease. To date, no screening tool has been widely adopted to identify patients with an increased risk of upstaging to invasive cancer, however, recent studies have evaluated algorithms and machine learning to identify this cohort of individuals. Jakub et al. recently developed a validated algorithm to determine the likelihood to be upstaged to invasive cancer, based on DCIS grade, presence of a mass, and multicentric disease.^12,22 Similarly, Hashiba et al. found the use of machine learning models including the following variables: palpable area of concern on presentation, mass visible on imaging, and grade 2 or 3 DCIS, could accurately predict individuals with DCIS that may be eligible for active surveillance.²³ In addition to preoperative risk stratification, the recent development of long-acting tracers for axillary mapping, such as superparamagnetic iron oxide (SPIO) nanoparticles, may be used in this setting to prevent the need for an axillary lymph node dissection in the event invasive disease is found after mastectomy.²⁴ SPIO nanoparticles are injected like other standard tracers and can last in the axilla for up to 248 days²⁵, thus allowing SLNB to performed at a subsequent operation in the event that an incidental invasive cancer is found on surgical pathology. The recent multicenter prospective SentiNot trial evaluated the use of SPIO nanoparticles in patients with DCIS undergoing mastectomy and 79% of patients in the study avoided surgical axillary staging.^12,25 The use of SPIO nanoparticles can also be used to address the dilemma in modern breast surgery where it is unknown if sentinel node should be utilized in oncoplastic breast surgery for large volume DCIS. Rearrangement of the tissue may prohibit future accurate sentinel node identification if invasive cancer is found on final pathology in these procedures.

Limitations to this study include those inherent to the NCDB such as its retrospective construct and manual data entry resulting in possible data entry error. Additionally, the NCDB has sparse data on physical exam findings such as palpable lesions which may have impacted nodal surgical decision-making and has been established as a risk factor for upgrade to invasive disease. Heterogeneity in the methods of preoperative tissue sampling, such as needle gauge size and number of passes, vary amongst institutions and may have resulted in variation in initial diagnostic accuracy. Despite these limitations, this study provides valuable insight into the high utilization of SLNB in patients undergoing mastectomy for DCIS and identifies a subset of patients in whom it may be reasonable to omit axillary nodal staging.

Conclusion

Most patients undergoing mastectomy for DCIS are currently undergoing surgical axillary staging in concordance with NCCN guideline recommendations. Given the low rate of nodal involvement (3.4%), it is reasonable to consider omitting SLNB for patients with low grade, hormone-receptor positive DCIS undergoing mastectomy. This data supports the role of future prospective investigations to determine the real-world long-term outcomes of omitting surgical axillary staging in select mastectomy patients with DCIS.

Clinical Practice Points.

The National Comprehensive Cancer Network guidelines do not recommend the use of sentinel lymph node biopsy (SLNB) for the treatment of ductal carcinoma in situ (DCIS), except in the setting of a mastectomy. The extent of utilization of SLNB in patients undergoing mastectomy for DCIS and factors associated with SLNB are unknown. The objectives of this study were to evaluate national SLNB utilization in women undergoing upfront mastectomy for DCIS, identify predictors of SLNB utilization, and determine the percentage with a positive SLNB. We utilized data from the National Cancer Database from 2012–2017. A total of 34,231 (88%) patients underwent SLNB and 4,742 (12%) had no surgical axillary staging. On multivariate analysis, younger age, high grade, hormone-receptor negative disease, and treatment at an academic institution were associated with an increased likelihood of SLNB utilization compared to their counterparts. We found 88% of patients underwent SLNB, consistent with national guidelines. Among them, only 3.4% were found to have nodal involvement. Predictors of a positive sentinel node in this population were non-Hispanic Black race and estrogen receptor negative disease. This study is the first to evaluate SLNB utilization in recent years and demonstrates low nodal involvement in patients with clinical DCIS undergoing mastectomy. Given the low rate of nodal involvement and associated complications of SLNB, it is reasonable to omit SLNB for the treatment of DCIS for patients with low grade, hormone-receptor positive DCIS. Future prospective studies are needed to determine the long-term implications of omitting SLNB in this population.

Highlights.

• Younger age and higher-grade disease are associated with SLNB utilization.

• Low nodal positivity was found in patients with DCIS undergoing mastectomy.

• Lower odds of a positive node on SLNB with estrogen receptor positive disease.

• Given low nodal positivity, omission of SLNB may be warranted in select population.