Update on Sentinel Node Pathology in Breast Cancer

Over the last three decades, the clinical management of the axilla in patients with breast cancer has undergone a paradigm shift towards de-escalating surgery to prevent overtreatment of patients. Sentinel lymph node biopsy (SLNB) has largely replaced axillary lymph node dissection (ALND) in clinically node-negative patients1. Practice-changing clinical trials have also shown that ALND can be safely avoided in early-stage breast cancer patients with limited involvement of the sentinel lymph nodes (SLNs)2, 3, 4. Evidence suggests that small metastases within the SLNs are of minimal clinical significance in this setting5,6. Although patients treated with neoadjuvant therapy were not included in these trials, the safety of SLNB in those who are node-negative7,8, or node-positive converted to node-negative9, 10, 11, has also recently been demonstrated.

Best practices in the pathologic evaluation of sentinel lymph nodes (SLNs) have been significantly impacted by these changes. While the value of intraoperative diagnosis of SLNs and of enhanced protocols to detect small metastatic deposits has been questioned in early-stage breast cancer patients, the optimal approach to SLN evaluation in patients treated with neoadjuvant therapy warrants special consideration. Nevertheless, practice patterns among pathologists remain variable with respect to the method of intraoperative examination, the approach to gross processing, and the utilization of hematoxylin and eosin (H&E) levels and cytokeratin immunohistochemistry (IHC)12. Herein, this review will provide evidence-based recommendations for SLN evaluation.

The SLNB procedure was validated in single-institution and multicenter studies13 as well as 7 prospective randomized controlled trials (RCTs)14, 15, 16, 17, 18, 19, 20, which confirmed that it is a safe and accurate method of staging in clinically node-negative patients. According to a meta-analysis of 69 studies of SLNB validated with concurrent ALND, SLNs are identified in >95% of cases and the average false negative rate (FNR) is <10%13. In the largest of the RCTs, the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-32 trial, there was no difference in local-regional recurrence (LRR), disease-free survival (DFS), or overall survival (OS) at 10-year follow up between patients who had SLNB alone versus ALND21. This and other trials also demonstrated that SLNB alone was associated with reduced arm morbidity, shorter recovery time, and better quality of life18, 22, 23, 24.

Practice-changing clinical trials have indicated that ALND can also be omitted in patients who have low-volume metastatic disease within the SLNs. Large, prospective RCTs investigated the noninferiority of no further axillary surgery to completion ALND in early-stage (cT1-2 N0) breast cancer patients with 1-2 positive SLNs. The American College of Surgeons Oncology Group (ACOSOG) Z0011 trial2 focused on patients with metastatic disease of any size involving one or two SLNs (n = 813). Other inclusion criteria included treatment with breast conserving therapy (BCT) with whole-breast radiotherapy as well as the absence of palpable adenopathy and gross extranodal disease; patients treated with neoadjuvant therapy were excluded. At a median follow-up of 6.3 years, there was no statistically significant difference in LRR or survival between the two arms of the study, and the comparable outcomes persisted at 10-year follow-up. The cumulative incidence of nodal recurrences at 10 years was 1.5% in the SLNB alone arm versus 0.5% in the ALND arm (p = 0.28)25. The International Breast Cancer Study Group (IBCSG) 23-01 trial3, which studied the significance of deposits no larger than a micrometastasis (≤ 0.2 cm) in patients who received BCT or mastectomy (n = 964) reported remarkably similar results when published in 2013 and at 10-year follow-up26 (Table 1).

The European Organization for Research and Treatment of Cancer (EORTC) After Mapping of the Axilla: Radiotherapy or Surgery? (AMAROS) trial4 investigated completion ALND versus axillary radiotherapy in patients with early-stage breast cancer (cT1-2 N0) and at least one positive SLN (n = 4806). At 10-year follow-up, there was a low rate of axillary recurrence in both treatment arms (1.82% in the axillary radiotherapy group vs. 0.93% in the completion ALND group) and no significant difference in distant metastasis-free survival (78.2% vs. 81.7%, respectively) or OS (81.4% vs. 84.6%, respectively)27 (Table 1). Axillary radiotherapy was associated with significantly less lymphedema as compared to ALND (11% vs 23% at 5 years, respectively; p<0.0001)4. The AMAROS trial results suggest that excellent axillary control is achieved with both axillary radiotherapy and ALND, but that radiotherapy is preferable due to the lower incidence of lymphedema. It is worth noting that in the Z0011 and the AMAROS trials additional positive axillary lymph nodes were identified in approximately 30% of patients who underwent completion ALND2,4. Although it may be difficult to accept that disease is left behind in so many SLNB patients, these trials underscore the importance of adjuvant radiotherapy and systemic therapy for adequate local-regional control and long-term outcomes.

American Society of Clinical Oncology (ASCO) clinical practice guidelines updated in 2016 recommend against ALND for women who have early-stage breast cancer (cT1-2) and one or two SLN metastases and who will undergo BCT with whole-breast radiotherapy28,29. Women who do not fall in this eligibility group include those with more than two positive nodes, gross extranodal extension (ENE), or palpable or matted axillary metastatic SLNs; women treated with neoadjuvant chemotherapy; and women treated with mastectomy. The current National Comprehensive Cancer Network (NCCN) guidelines30 are consistent with the ones published by ASCO, except omission of ALND is explicitly allowed in mastectomy patients with SLN micrometastases.

With the introduction of the SLNB procedure, the SLN was subjected to extensive examination with step sectioning and cytokeratin IHC, leading to an increased detection of small metastases31, 32, 33, 34, 35. The distinction between isolated tumor cells (ITCs; < 0.2 mm and < 200 cells) and micrometastases (0.2 mm to 2 mm) was originally incorporated into the 6th edition of the American Joint Committee on Carcinoma (AJCC) Tumor Node Metastasis (TNM) staging system in 200236, and persists today in the 8th edition37. Enhanced pathology protocols for SLN examination have been pursued with the belief that the identification of occult metastases, or small metastases not seen on the initial hematoxylin and eosin (H&E) slide but detected on deeper H&E or IHC levels, may predict for additional non-sentinel axillary lymph node involvement and axillary recurrence. In the post-Z0011 era, the significance of these small metastases in early-stage breast cancer patients has been questioned37,38, given that there is no benefit to ALND in patients who have up to 2 positive nodes, and this is true despite leaving disease behind in patients treated with SLNB alone.

Numerous retrospective studies have investigated the significance of low volume or occult disease in the axillary lymph nodes with conflicting results. Meta-analyses of the early literature showed that the pooled risk of additional metastases in non-SLNs was 20% for micrometastases, 12.3% for ITCs, and 9% for occult metastases detected by IHC only39,40. A population-based study of women with early-stage breast cancer using Surveillance, Epidemiology and End Results (SEER) data found that occult micrometastases detected via IHC were an independent predictor of worse survival, whereas occult ITCs were of no prognostic significance41. A retrospective study of women with early-stage breast cancer the Netherlands Cancer Registry by de Boer et al. demonstrated that both ITCs and micrometastases were associated with reduced 5-year DFS, but this prognostic effect was erased in patients who received adjuvant chemotherapy42. In a systematic review of the literature prior to 2010 conducted by the same authors, both the presence of metastases ≤2 mm and the presence of occult metastases had a detrimental effect on survival43. A significant limitation is that many of the included studies were conducted prior to the introduction of SLNB and the modern era of breast cancer care.

Contemporary prospective studies with standardized protocols for SLN examination have not supported the use of enhanced protocols to detect occult metastases. The NSABP B-32 trial reported on the clinical impact of occult metastases in patients who were SLN-negative on an initial H&E section5. Initial pathologic examination of the SLNs involved slicing them at 2 mm intervals, submitting them entirely, and examining a single H&E section. Additional sections were widely spaced at 0.5 mm and 1.0 mm deeper in the block, with H&E and IHC performed at each level. Occult metastases were identified on deeper levels in 16% of patients, of which 72% were ITCs, 28% were micrometastases, and only 0.02% were macrometastases (>2 mm). Five and 10-year follow-up of patients with and without occult metastases revealed statistically significant differences in DFS and OS, but the magnitude of the percent difference was minimal (DFS: 86.4% vs. 89.2%, OS: 94.6% vs. 95.8%), suggesting that there is no clinical benefit to finding these overlooked metastases5,44. Likewise, the ACOSOG Z0010 study, a prospective observational study of occult metastases detected in one deeper level used for cytokeratin IHC in 3,326 SLN specimens, found no significant difference in the 5-year DFS or OS between patients with and without occult metastases6. Similar conclusions can be drawn from the results of both studies: multiple levels and cytokeratin IHC are unnecessary.

Recommendations from the College of American Pathologists (CAP)38 and the American Joint Committee on Cancer (AJCC)37 agree that the priority of SLN evaluation is to detect all macrometastases (>2 mm). As such, the SLNs should be thinly sliced in 2 mm increments, submitted entirely, and evaluated by at least a single H&E level. Although it is acknowledged that multilevel sectioning and IHC may reveal additional micrometastases and ITCs, the clinical impact of these small metastatic deposits is minimal.

Metastasis of invasive lobular carcinoma (ILC) to the lymph nodes warrants special consideration due to its pattern of spread as noncohesive tumor cells. It has been debated whether the results of the Z0011 trial are applicable to patients with lobular histology given that they represented only 7% of trial participants45. In a recent study of Z0011-eligible patients, patients with ILC more often had SLN macrometastases and >2 mm of ENE, but the proportion of patients with 3 or more positive SLNs did not significantly differ by histologic subtype, suggesting that ALND is not warranted on the basis of histology46. Conflicting evidence exists regarding whether lobular histology is associated with a higher risk of non-SLN involvement47, 48, 49, 50, 51, yet several studies have revealed no significant difference in the overall rate of nodal positivity of ILC versus invasive ductal carcinoma52, 53, 54, 55, 56. Early-stage breast cancer patients with ILC are more likely to have ITCs and micrometastases in the axillary SLNs57, 58, 59, but the clinical significance of this finding remains uncertain.

The interpretation of SLNs in patients with classic ILC is notoriously difficult due to the dispersed single-cell pattern of involvement and the frequently bland cytologic features. This can lead to missing nodal metastases, even relatively large ones60, on frozen section or routine H&E analysis alone. Lymph nodes from patients with ILC can be found to harbor occult metastases detected on cytokeratin IHC in up to 40% of patients61,62. Therefore, the routine use of cytokeratin IHC is recommended for SLN evaluation in cases with lobular morphology, as this will not only improve the detection of metastatic disease, but also highlight its extent. Because ILC metastasizes as individual cells and not as cohesive clusters, it can be difficult to determine the appropriate size of the largest metastatic deposit. In cases with a dispersed pattern of nodal involvement, it is recommended that the pathologist use his or her judgment to assign the best N category and to include the reason for the difficulty in classification in a note38(Fig 1a-1d).

The incidence of ENE in the SLN, defined as the extension of metastatic cells through the nodal capsule into the perinodal adipose tissue63, is 24-40%63, 64, 65, 66. Gross ENE was an exclusion criterion in the Z0011 trial2, but the significance of microscopic ENE was not investigated. Prior studies have demonstrated that ENE associated with axillary SLN metastasis is an important predictor of non-sentinel axillary nodal burden, disease recurrence, and overall mortality40,63,65.

Studies have also investigated the significance of ENE specifically in patients who would have been eligible for the Z0011 trial. In a recent analysis of such patients by Gooch et al.67, 33% of patients with > 2 mm of ENE had four or more positive nodes, whereas only 9% of patients with ≤ 2 mm of ENE had four or more positive nodes, leading the authors to argue that ALND should be recommended for > 2 mm of ENE67. Choi et al.68 similarly found that the extent of ENE was associated with the mean number of non-sentinel lymph node metastases (0.48 for no ENE vs. 1.91 for ≤ 2 mm of ENE vs. 2.95 for > 2 mm of ENE). There were no significant differences in recurrence or survival in those with and without ENE. Barrio et al. found no difference in nodal recurrence rates in patients with and without microscopic ENE, with a low overall rate at 5 years of 1.6%, and suggested that the presence of microscopic ENE should not be considered an indication for ALND69.

It is advisable to report not only the presence of ENE, a CAP-recommended element38, but also the extent of ENE68(Figure 2a-2d), which may predict for non-SLN involvement and may influence the decision to perform ALND. In the absence of an accepted standard for measurement of ENE in the CAP or AJCC guidelines37,38, the authors of the aforementioned studies proposed measurement of the widest diameter of the invasive front of ENE, either perpendicular or parallel to the nodal capsule68,69. This method, the measurement of the maximal dimension of ENE regardless of the relationship to the capsule, is one that we have employed in our practice and that other pathologists have recommended70. It is worth noting that other recent studies have investigated the prognostic significance of the perpendicular and circumferential diameters of ENE71,72, with one study finding that the perpendicular diameter was the only measurement to predict for disease-free and breast cancer specific survival72. Further data is needed for evidence-based recommendations.

The use of SLNB in patients with ductal carcinoma in situ (DCIS) is controversial due to the low risk of axillary metastases, even when upstaged to invasive cancer at final pathology, with rates ranging from 4% to 15%1. ASCO and NCCN guidelines discourage the use of SLNB in patients with DCIS treated with BCT unless the location of the surgery may compromise future SLN mapping or the clinical findings suggest a high risk of upgrade to invasive cancer, such as the presence of a mass on clinical breast exam or imaging or calcifications spanning more than 5 cm29,30. In contrast, SLNB should be strongly considered in those who undergo mastectomy, because this surgery precludes subsequent SNB at a second operation29,30, and furthermore, the rate of upstaging to invasive cancer may be as high as 28% to 48% in this population73. The approach to pathologic evaluation of SLNs in patients with early-stage invasive breast cancer can be extrapolated to those with DCIS undergoing mastectomy.

Neoadjuvant chemotherapy (NACT), or preoperative chemotherapy, is the standard of care for locally advanced and inflammatory breast cancer and has been increasingly employed in the management of patients with earlier stage disease74. The advantages of NACT include the potential for breast conservation and less extensive axillary surgery, the assessment of response to therapy, and the provision of prognostic information based on the pathologic response. A pathologic complete response (pCR) in the axilla is seen in approximately 40% of patients75, which suggests that NACT has the potential to spare many patients from ALND.

SLNB has been established in clinically node-negative patients treated with NACT. Two meta-analyses have revealed pooled SLN identification rates (89% to 91%) and FNRs (8.4% to 12%) that are comparable to those reported for patients undergoing upfront surgery7,8. An increasing body of evidence also suggests that selected patients with biopsy-proven nodal involvement may be candidates for SLNB if they downstage to a clinically negative axilla. Three large prospective RCTs have investigated the feasibility and accuracy of SLNB in patients who had biopsy-proven axillary metastases prior to NACT: the ACOSOG Z1071 trial9, the Sentinel Lymph Node Biopsy in Patients with Breast Cancer before and after Neoadjuvant Chemotherapy (SENTINA) trial10, and the Sentinel Node Biopsy Following Neoadjuvant Chemotherapy in Biopsy Proven Node Positive Breast Cancer (SN FNAC) trial11.

In the ACOSOG Z1071 trial9 and Arm C of the SENTINA trial10, patients with clinically node-positive breast cancer underwent SLNB with immediate ALND following NACT. Both trials revealed an FNR > 10% (12.6% in Z1071 and 14.2% in SENTINA), which exceeded the pre-specified acceptable threshold set by the Z1071 trial. Although the high FNR raised concern regarding the safety of SLNB in the neoadjuvant setting, both trials also demonstrated that the FNR was significantly decreased with the use of dual-tracer mapping (10.8% in Z1071, 8.6% in SENTINA) and with the removal of at least 3 SLNs (9.1% in Z1071, 4.9% in SENTINA). In the SN FNAC trial11, unlike the other trials, IHC analysis of the SLNs was mandated, and SLNs involved by metastases of any size, including ITCs, were classified as positive. The FNR was 8.4%; however, the FNR increased to 13.3% if nodes with tumor cells detected only by IHC were considered negative(Table 2). Thus, it can be argued that SLNB is a safe alternative to ALND when dual tracers are used and at least 2 or 3 SLNs are identified, and that IHC may be of value in the analysis of SLNs in the neoadjuvant setting.

In a subsequent analysis of the Z1071 trial data, the identification and resection of the biopsy-proven positive lymph node marked with a metallic clip prior to NACT decreased the FNR of SLNB to 6.8%76. Other studies have demonstrated that wire or radioactive seed localization significantly improves the success rate of surgical removal of clipped nodes and the accuracy of SLNB after neoadjuvant therapy77,78. Targeted axillary dissection, which was first described as a procedure involving preoperative localization of the clipped node with iodine-125 seeds followed by retrieval of the clipped node along with the SLN(s), yielded an FNR as low as 1.4%79. Other methods to mark lymph nodes include ferromagnetic seeds, infrared-activated reflectors, and carbon tattooing80,81. Marked lymph node biopsy and targeted axillary lymph node dissection has been incorporated in the NCCN guidelines30.

Recurrence and survival data awaits ongoing clinical trials. The NSABP B51 trial is investigating regional nodal irradiation versus no treatment in patients who convert to node-negative disease following NACT82. The Alliance for Clinical Trials in Oncology A11202 trial is currently assessing the effectiveness of axillary radiation as an alternative to ALND for patients with clinically node-positive breast cancer who become clinically node-negative following NACT83. Until the publication of these results, ALND remains the standard of care for patients with residual disease in the SLNs after NACT30.

Pathologic assessment of the response to neoadjuvant therapy and the amount of residual disease in the sentinel and non-sentinel axillary lymph nodes provides important prognostic information. Patients who achieve axillary pCR have significantly improved 5-year recurrence-free survival and OS compared with those who have residual nodal disease84. In patients who have histologic evidence of regression in the lymph nodes, the survival is intermediate between those with residual nodal disease and those with negative nodes85. Survival decreases with an increasing number of involved lymph nodes84, 86, 87, 88 and with increasing size of the residual nodal disease89,90.

The significance of microscopic nodal metastases in patients treated with NACT is controversial. Moo et al.91 observed that residual micrometastases and ITCs in the SLNs predict for higher rates of positive non-SLNs at ALND than seen in the primary surgery setting. The likelihood of positive non-SLNs was similar among patients with SLN micrometastases and macrometastases, 64% versus 62%, respectively. Klauber-DeMore et al.89 demonstrated that residual micrometastases (<2 mm) were associated with worse DDFS and OS than negative lymph nodes, and the outcomes were comparable to those observed for macrometastases of 2 mm to 2 cm in size. Similarly, Wong et al.90 reported significantly worse 5-year DFS in patients with residual ITCs and residual micrometastases compared to those who had an axillary pCR (73.5% and 74.7% versus 88.4%, p < .001). Using NCDB data from 35,536 patients, the same authors found that 5-year OS rates were significantly worse with increasing nodal disease burden (ypN0, 88.9%; ypN0[i+], 82.8%, ypN1mi, 79.5%; and ypN1, 77.6%; p < 0.001). In contrast, a retrospective analysis of data from the Netherlands Cancer Registry suggested that DFS and OS were comparable between ypN0(i+)/ypN1mi and ypN0 disease92. In the absence of outcomes data from clinical trials, the potential prognostic impact of residual low-volume nodal disease cannot be dismissed.

Given the prognostic significance of the post-treatment nodal status, pathologic evaluation of SLNB and ALND specimens should be undertaken with great care and with an appreciation of special considerations. The pathology report should include the number of involved lymph nodes, the size of the largest metastatic deposit, the presence of ENE, the presence of treatment effect in both involved and uninvolved nodes, and the presence of biopsy site changes. The AJCC TNM37 and the Residual Cancer Burden (RCB)93 systems are most commonly employed for pathologic assessment of response to therapy, and the number of involved lymph nodes and the size of the largest residual metastatic deposit are integral to both. Post-treatment metastatic deposits are often characterized by single cells and small clusters of cells scattered within histologic changes of therapy-induced regression, such as fibrosis, histiocytic inflammation, hemosiderin deposition, dystrophic calcifications, and occasionally pools of mucin. In such cases, determining the appropriate size and stage is not always straightforward.

The AJCC 8th edition was updated to clarify that the ypN stage is determined by the size of the largest contiguous focus of viable metastatic carcinoma excluding intervening treatment-related fibrosis37(Fig 3a-3h). In contrast, recommendations from an international working group published prior to the AJCC 8th edition suggested that the size of the entire area, including tumor cells and intervening fibrotic stroma, should be measured rather than the size of the largest cell cluster94. With strict adherence to the AJCC guidelines, it must be acknowledged that some deposits with treatment effect may be classified as ITCs when they likely represent downstaging of lymph nodes with previous involvement by micro- or macrometastatic disease, and this finding may have different biologic implications than de novo ITCs. Although lymph nodes with ITCs after chemotherapy are staged as ypN0(i+), the AJCC recommends against regarding this as an axillary pCR37, which may be diagnosed only in the absence of any viable tumor cells in the lymph nodes. In our practice, we find it can be helpful to include a description of the findings in an explanatory note for challenging cases requiring judgement.

It remains uncertain whether enhanced pathology protocols for SLN evaluation, such as multiple H&E levels and/or cytokeratin IHC, should be routinely applied in the neoadjuvant setting. Although one prior study found no statistically significant difference in DFS or OS between patients with and without occult metastases on two additional levels for H&E and cytokeratin IHC staining95, conclusions are limited by the small size of this study and the evidence that small metastases may be of prognostic significance. As the use of cytokeratin IHC decreases the FNR of SLNB procedure in the neoadjuvant setting, the American Society of Breast Surgeons recommends that IHC staining of the SLNs should be considered, particularly in patients with node-positive disease at presentation96. Practically speaking, cytokeratin IHC is often useful in highlighting tumor cells within treatment related changes and confirming the true burden of residual disease, and as such, a low threshold for performing cytokeratin IHC in the post-neoadjuvant setting is prudent.

In the post-Z0011 era, there has been a progressive decline in the use of intraoperative evaluation of axillary SLNs due to the more limited circumstances in which ALND is performed97, 98, 99, 100. In one single institution study, there was a dramatic drop in frozen section analysis of axillary SLNs from 69% to 2% of the cases100. SLNs from clinically node negative patients who would have met Z0011 and/or AMAROS trial criteria can be sent directly for permanent processing, as only a small fraction of patients will be found to have 3 or more positive SLNs requiring completion ALND101. Because mastectomy patients were excluded from the Z0011 trial2 and were not well-represented in the 23-01 (9% of cases)3 and AMAROS (17% of cases)4 cohorts, controversy remains regarding axillary management in mastectomy patients, and the utilization of intraoperative evaluation of SLNs may depend on the clinical situation and institutional practice patterns102, 103, 104. Intraoperative evaluation of SLNs is still of clear value in patients treated with NACT who are clinically node-negative at the time of surgery. In this population, ALND remains the standard of care for any volume of disease96, and as such, the intraoperative detection of even small metastases may be of great clinical significance.

In the processing of SLNs for intraoperative consultation, every attempt should be made to identify, dissect, and enumerate individual lymph nodes from packets of tissue, which may require trimming the fat. It is especially important to provide the surgeon with an accurate lymph node count in post-treatment cases where the removal of at least 2 or 3 SLNs is required for an acceptable FNR9, 10, 11 and anything less may result in ALND96. The presence of a clipped node should be confirmed, and any localizing markers, such as seeds or reflectors, must be retrieved. Failure to remove the initially biopsy-proven positive lymph node may also lead to ALND96. SLNs should be sectioned at 2 mm intervals and those without grossly evident tumor should be submitted entirely for histologic evaluation.

Several methods may be employed in the intraoperative assessment of SLNs. Frozen section analysis, the most common method of evaluation12, has an overall sensitivity of 78%, with higher sensitivity for macrometastasis than micrometastasis (94% vs. 40%)105. Most false-negative results are due to sampling error (up to 94% of cases)106, 107, 108, 109, with no tumor cells apparent on the frozen section slides. This underscores the importance of careful serial sectioning of lymph nodes at 2 mm intervals and histologic sectioning of frozen tissue to optimize the surface area of the tissue represented on the frozen slides108,109. Size of the metastatic disease is one of the most important histologic factors influencing the FNR107, 108, 109, 110, 111, 112, 113, 114, 115, 116, with up to an 8-fold increase in FNR reported for micrometastatic disease compared to macrometastatic disease110. Interpretative errors in SLN frozen section analysis are associated with tumor characteristics such as low-grade cytology106, 107, 108,117 and lobular histologic subtype110,112,115,118. Interpretation of the SLN frozen section following NACT may also be especially challenging due to the reduced cellularity of metastatic deposits and the obscuring effect of treatment-related changes. Despite this, Grabbenstetter et al.108 recently demonstrated that frozen section analysis is a reliable method of SLN assessment in the neoadjuvant setting, with an FNR of only 5.4%.

It has been argued that intraoperative cytology – including touch imprint, scrape, and smear techniques - is less expensive, quicker, easier to perform, and better at preserving tissue than frozen section. However, touch imprint cytology has a lower sensitivity than frozen section, with an overall sensitivity of 63%119. A significant limitation is that cytologic techniques cannot precisely determine the size of a metastasis (Fig 4a-4d). This may be problematic in certain clinical circumstances such as the consideration of conservative management for low-volume disease in a patient who would not have met Z0011 criteria. Intraoperative cytology may be best reserved as an adjunct to frozen section rather than as a standalone approach.

It has been suggested that rapid IHC and molecular techniques may improve the accuracy of intraoperative SLN evaluation. Rapid cytokeratin IHC, when compared to frozen section and touch imprint cytology, is the least sensitive method for detecting SLN metastases120, yet when used in combination with frozen section, the sensitivity of intraoperative SLN evaluation improves to a level comparable to final pathology121. One-Step Nucleic Acid (OSNA), which is a molecular technique developed by Tsujimoto and colleagues122, employs reverse transcription, loop-mediated isothermal amplification of cytokeratin 19 mRNA to classify SLNs as negative, micrometastasis, or macrometastasis. OSNA has a high sensitivity (87%) for detecting macrometastases; however, over 20% of patients diagnosed with macrometastases by OSNA would have disease re-classified as micrometastases on histopathologic examination123. The potential for discrepant results at an important nodal staging threshold is a major criticism of this method. False negatives are possible with cytokeratin 19-negative tumors, while rare false positives may be seen with benign inclusions. Neither rapid IHC nor OSNA has not gained widespread use in the United States.

Benign findings in the lymph nodes may be misinterpreted as metastatic carcinoma. A false-positive diagnosis, whether rendered at the time of intraoperative consultation or final review, may result in unnecessary ALND.

Nevus cell rests, which consist of small clusters of benign melanocytic cells within the capsule (Fig 5a), are a relatively common finding in the axillary lymph nodes124. Although the morphologic distinction of capsular melanocytic nevi from metastatic breast carcinoma is usually straightforward, the diagnosis is supported by the presence of staining for melanocytic markers, like Melan A (Fig 5b), and the absence of staining for cytokeratin on IHC studies125.

Benign glandular inclusions within the axillary lymph nodes may be mammary-type or Mullerian-type. Heterotopic breast tissue consists of ductal and lobular elements with retention of the normal inner luminal epithelial and outer myoepithelial layers, with the latter highlighted by IHC for myoepithelial markers. The lesion may demonstrate the spectrum of non-proliferative and proliferative changes seen in the breast, including apocrine metaplasia, squamous metaplasia, cysts, epithelial hyperplasia, adenosis, and papillary proliferations, and rarely may be involved by atypical hyperplasia or carcinoma in situ125,126(Fig 5c-5e). Endosalpingiosis, which is characterized by ciliated glandular epithelium resembling that of the fallopian tube (Fig 5f), is quite rare in the axillary lymph nodes, and can be confirmed by IHC staining for the Mullerian marker PAX8125,127.

Benign breast epithelium may also be found in axillary lymph nodes as a result of iatrogenic displacement and mechanical transport, most frequently following core biopsy of papillary lesions. Findings in the lymph node that argue for benign transport include single cells or papillary clusters in the sinuses, associated foamy histiocytes and foreign body giant cells, as well as bland cytology and/or IHC profile different than the primary tumor128.

The application of artificial intelligence (AI) to histologic assessment of lymph node metastases has generated significant interest. Bejnordi et al.129 assessed the performance of automated deep learning algorithms at detecting metastases in tissue slides from breast cancer patients. The top algorithm performed significantly better on a whole-slide image classification task than a panel of 11 pathologists in diagnostic simulation with time-constraints. Holten-Rossing et al.130 investigated whether digital image analysis by the AI could be used as a screening tool for SNLB slide assessment. The algorithm, which was applied to SLNBs from 135 patients, had a sensitivity of 100%. The authors argued that the algorithm could decrease the workload for pathologists by over 50%.

Although the technical capabilities are exciting, the utility of AI algorithms in clinical practice remains to be determined. Since the intraoperative examination of SLNs in patients with breast cancer has dramatically decreased in recent years, the opportunity to employ such algorithms in this context is limited. Also, while AI may offer modest improvement in detecting small metastases, this may be of limited clinical value in this era of conservative management of the axilla.

留言 (0)

沒有登入
gif