Introduction
The Choosing Wisely recommendations advocate against routine use of sentinel lymph node biopsy (SLNB) for women ≥70 with estrogen receptor-positive (ER+), clinically node-negative (cN0) breast cancer (BC). However, de-escalation of SLNB remains challenging, perhaps because the broad recommendation does not consider subgroups where SLNB may remain beneficial1, 2. One such subgroup may be invasive lobular carcinoma (ILC), which is associated with more difficult detection on screening imaging, larger size and multifocality at diagnosis, and more nodal involvement3.
We sought to determine if older women with ILC who do not undergo SLNB experience increased proportions of recurrences and a concomitant decrease in locoregional recurrence free survival (LRFS) compared to women with invasive ductal carcinoma (IDC).
Methods
This study included women with stage I-II ER+, HER2−, cN0 BC aged ≥70 years treated between 2010-8 within a healthcare system that included 15 community and academic hospitals. This study utilized linked cancer registry and electronic medical record data4. The University of Pittsburgh Institutional Review Board approved this study (IRB19010319).
The primary outcome was LRFS. LRFS analysis was performed using a conditional landmark time to reduce immortal time bias5 and was thus defined as the time from 60 days after diagnosis (around the average time from diagnosis to surgery and recovery) to a local or regional recurrence (as previously defined in ref. 4), or to censoring if lost to follow up (not including second primary cancers). Non-BC-specific mortality events occurring before a recurrence were considered censored events. Median follow up was 6.5 years. Patients with ILC who did not undergo SLNB were compared to patients with ILC who did undergo SLNB as well as to patients with IDC who did not undergo SLNB. Kaplan-Meier curves with log-rank test compared LRFS. P-values <0.05 were considered statistically significant.
Results
Our cohort included 1261 patients with IDC and 251 with ILC. Patients with ILC had larger tumors (p<0.0001) and higher stages at diagnosis (p<0.0001) (Table 1).
TABLE 1:
Characteristics of patients with IDC and ILC included in the study.
| Patients with IDC (n = 1261) | Patients with ILC (n = 251) | P-value | ||
|---|---|---|---|---|
| Median age at diagnosis, years (IQR) | 76 (72-81) | 76 (73-81) | 0.41 | |
| Median tumor size, mm (IQR) | 13 (8-19) | 17.5 (12-25) | < 0.0001 | |
| Number of Comorbidities | 0.02 | |||
| 0 | 524 (41.6%) | 127 (50.6%) | ||
| 1 | 209 (16.6%) | 30 (12.0%) | ||
| ≥ 2 | 85 (6.7%) | 12 (4.8%) | ||
| Unknown | 443 (35.1%) | 82 (32.7%) | ||
| Grade at Diagnosis | 0.0003 | |||
| I | 299 (23.7%) | 41 (16.3%) | ||
| II | 715 (56.7%) | 166 (66.1%) | ||
| III | 204 (16.2%) | 22 (8.8%) | ||
| Unknown | 43 (3.4%) | 22 (8.8%) | ||
| Clinical Stage at Diagnosis | < 0.0001 | |||
| I | 971 (77.0%) | 145 (57.8%) | ||
| II | 290 (23.0%) | 106 (42.2%) | ||
| Underwent SLNB | 928 (73.6%) | 188 (74.9%) | 0.69 | |
| Pathologic Node Positive After SLNB | 98 (10.6%) | 12 (6.4%) | 0.08 | |
| Total Recurrences | ||||
| After Omission of SLNB | 15 / 333 (4.5%) | 4 / 63 (6.3%) ^ | 0.52 | |
| After Undergoing SLNB | 46 / 928 (5.0%) | 7 / 188 (3.7%) ^ | 0.57 | |
| Locoregional Recurrences * | ||||
| After Omission of SLNB | 5 / 333 (1.5%) [3 in-breast, 2 axilla] |
1 / 63 (1.6%) # [1 in-breast] |
0.99 | |
| After Undergoing SLNB | 23 / 928 (2.5%) [14 in-breast, 5 axilla, 2 chest wall, 2 skin over breast] |
4 / 188 (2.1%) # [2 in-breast & axillary LN, 2 chest wall] |
0.99 | |
| Breast Surgery & | 0.06 | |||
| Breast Conservation | 907 (71.9%) | 166 (66.1%) | ||
| Mastectomy | 354 (28.1%) | 85 (33.9%) | ||
| Underwent Adjuvant Radiation Therapy | 798 (63.2%) | 153 (61.0%) | 0.52 | |
| Received Adjuvant Endocrine Therapy | 1118 (88.7%) | 223 (88.8%) | 0.99 | |
If comparing total recurrences between patients with ILC omitting SLNB versus patients with ILC undergoing SLNB, Fisher’s exact testing yields a p-value of 0.48, indicating a non-significant difference.
If comparing locoregional recurrences between patients with ILC omitting SLNB versus patients with ILC undergoing SLNB, Fisher’s exact testing yields a p-value of 0.99, indicating a non-significant difference.
Five-year axillary recurrence rates were below 1% for patients with IDC who did and did not undergo SLNB as well as for patients with ILC who did and did not undergo SLNB.
Rates of use of SLNB varied depending on the surgery type: for patients with IDC, 77.3% of patents underwent SLNB with breast conversation and 64.1% underwent SLNB with mastectomy. Similarly, in patients with ILC, 81.3% of patients underwent SLNB with breast conservation and 62.4% of patients underwent SLNB with mastectomy.
Patient characteristics were compared using Mann-Whitney, Fisher’s exact, or Chi-square tests.
Race and ethnicity data was not available for this cohort of patients.
Equivalent proportions of patients underwent SLNB (IDC: 73.6%, ILC: 74.9%; p=0.69). In patients undergoing SLNB, those with ILC (6.4%) harbored lower rates of pathologically positive SLNs compared to those with IDC (10.6%) (p=0.08). Patients with ILC who omitted SLNB did not experience a significant increase in proportions of total or locoregional recurrences compared to patients with IDC who omitted SLNB, and when compared to patients with ILC who underwent SLNB.
There was no difference in LRFS between patients with IDC who did and did not undergo SLNB (p=0.70) and between patients with ILC who did and did not undergo SLNB (p=0.49) (Figure 1A&1B). When comparing patients with IDC and with ILC who did not undergo SLNB, there was no difference in LRFS (p=0.61), nor a difference in LRFS in those who did (p=0.37) (Figure 1C&1D).
FIGURE 1:
Locoregional recurrence free survival (LRFS) between key comparison groups. (A) Comparison of LRFS between patients with IDC who did and did not undergo SLNB (p = 0.70). (B) Comparison of LRFS between patients with ILC who did and did not undergo SLNB (p = 0.49). (C) Comparison of LRFS between patients with ILC and patients with IDC who did not undergo SLNB (p = 0.61). (D) Comparison of LRFS between patients with ILC and patients with IDC who did undergo SLNB (p = 0.37).
Discussion
Surgeons have found it difficult to broadly apply the Choosing Wisely recommendation for omitting SLNB to all patients ≥70 with early-stage ER+ BC. An important subgroup for consideration is patients with ILC.
Limited data exist for older women with ILC regarding recurrence when omitting SLNB. Key references guiding these recommendations from the Society of Surgical Oncology fail to provide lobular-specific data in older women despite the prevalence and unique features of this tumor type6. In this study, with a comparably large and well-annotated cohort of patients with ILC, we found that patients ≥70 with ILC who omitted SLNB had neither increased recurrences nor inferior LRFS compared to those who did undergo SLNB and compared to patients with IDC. This is despite patients with ILC harboring larger tumors with accompanying higher stages.
Limitations include the sample size and follow up time. Future studies with larger sample sizes and increased power should be used to validate these findings, although in this population achieving power given the low rate of recurrence may be challenging. Ongoing clinical trials such as SOUND (NCT02167490) and INSEMA (NCT02466737) may help to shed light on omission of SLNB for patients with ILC. Despite these limitations, these data lend support to the notion that the Choosing Wisely recommendations can be extended to patients with ILC.
ACKNOWLEDGEMENTS
We acknowledge the many contributions of the patients, families, researchers, clinical staff, and sponsors of this study. The authors of this study thank the members of the Lee-Oesterreich lab for their helpful discussions and support. The Department of Clinical Analytics in the UPMC Health Services Division generated the raw data for this analysis with the support of the UPMC Network Cancer Registry.
Funding & Support:
Shear Family Foundation (to Adrian V. Lee & Steffi Oesterreich), National Cancer Institute under award number 1F30CA264963-01 (to Neil Carleton), and a UPMC Hillman Cancer Center Developmental Pilot Grant (to Adrian V. Lee & Steffi Oesterreich). Steffi Oesterreich and Adrian Lee are also funded by the Breast Cancer Research Foundation and are Komen Scholars. The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Footnotes
Conflicts of Interest and Disclosures: The authors have no conflicts of interest or other items to disclose.
REFERENCES
- 1.Park KU, Bazan JG. Implications of Omitting Sentinel Lymph Node Biopsy in Patients Older Than 70 Years. JAMA Surg. Feb 1 2021;156(2):199–200. doi: 10.1001/jamasurg.2020.5008 [DOI] [PubMed] [Google Scholar]
- 2.Wang T, Baskin AS, Dossett LA. Deimplementation of the Choosing Wisely Recommendations for Low-Value Breast Cancer Surgery: A Systematic Review. JAMA Surg. Aug 1 2020;155(8):759–770. doi: 10.1001/jamasurg.2020.0322 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Mukhtar RA, Brabham CE, Guo R, et al. Accuracy of sentinel lymph node biopsy in invasive lobular carcinoma of the breast: Factors associated with false negatives. Breast J. Apr 2021;27(4):406–408. doi: 10.1111/tbj.14161 [DOI] [PubMed] [Google Scholar]
- 4.Carleton N, Zou J, Fang Y, et al. Outcomes After Sentinel Lymph Node Biopsy and Radiotherapy in Older Women With Early-Stage, Estrogen Receptor-Positive Breast Cancer. JAMA Netw Open. Apr 1 2021;4(4):e216322. doi: 10.1001/jamanetworkopen.2021.6322 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Giobbie-Hurder A, Gelber RD, Regan MM. Challenges of guarantee-time bias. J Clin Oncol. Aug 10 2013;31(23):2963–9. doi: 10.1200/jco.2013.49.5283 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Chung A, Gangi A, Amersi F, Zhang X, Giuliano A. Not Performing a Sentinel Node Biopsy for Older Patients With Early-Stage Invasive Breast Cancer. JAMA Surgery. 2015;150(7):683–684. doi: 10.1001/jamasurg.2015.0647 [DOI] [PubMed] [Google Scholar]