- Email: [email protected]
Intraoperative sentinel lymph node evaluation: Optimizing surgical pathology practices in an era of changing clinical management
Annals of Diagnostic Pathology 33 (2018) 45–50
Contents lists available at ScienceDirect
Annals of Diagnostic Pathology journal homepage: www.elsevier.com/locate/anndiagpath
Intraoperative sentinel lymph node evaluation: Optimizing surgical pathology practices in an era of changing clinical management
T
⁎
Margaret L. Comptona, , Raeshell S. Sweetingb, Emily S. Reisenbichlera,c a
Department of Pathology, Microbiology and Immunology, Vanderbilt University Medical Center, 1161 21st Avenue South CC-3322 Medical Center North, Nashville, TN 37232-2561, United States b Department of Surgery, Vanderbilt University Medical Center, 1161 21st Avenue South CC-3322 Medical Center North, Nashville, TN 37232-2561, United States c Yale University, Department of Pathology, 310 Cedar Street, PO Box 208023, New Haven, CT 06520-8023, United States
A R T I C L E I N F O
A B S T R A C T
Keywords: Breast Axilla Sentinel lymph node Intraoperative evaluation Surgical pathology
Axillary lymph node status is an independent prognostic indicator in breast cancer. Intraoperative identification of metastatic carcinoma in sentinel lymph nodes may allow for concurrent axillary lymph node dissection at the time of primary tumor excision. A retrospective review of patients undergoing primary breast cancer excision with sentinel lymph node sampling was performed. Sensitivity and specificity of imprint cytology (touch prep) with and without the incorporation of gross evaluation was determined using permanent section results as the gold standard. Five hundred sixteen lymph nodes were analyzed by imprint cytology in 213 patients, and 203 lymph nodes were analyzed in 74 patients incorporating gross examination. Sensitivity and specificity for the detection of macrometastases by touch prep alone were 60% and 99% respectively with 4 patients undergoing same-day axillary dissection for only micrometastatic disease. False negative causes included lack of transfer of malignant cells in 8 cases and misinterpretation of tumor cells in 6 cases. Incorporating gross examination in the modified protocol resulted in reduced sensitivity of 38%, but achieved the desired 100% specificity and positive predictive value. Imprint cytology alone did not reliably distinguish between micro- and macrometastatic disease. Gross assessment combined with imprint cytology allows for improved assessment of volume of axillary disease, but is an insensitive technique.
1. Introduction Axillary lymph node status is an independent prognostic indicator in breast cancer. In patients with a clinically negative axilla, intraoperative identification of macrometastatic carcinoma (> 2 mm tumor deposit) in sentinel lymph nodes allows for concurrent axillary lymph node dissection (ALND) and complete pathologic staging at the time of primary tumor excision. However, ALND may be associated with morbidity including pain, nerve damage, loss of mobility, lymphedema and, per the American College of Surgeons Oncology Group (ACOSOG) Z0011 clinical trial and the International Breast Cancer Study Group 23-01 clinical trial, has not been shown to improve overall survival in selected groups of women with isolated metastatic disease [1]. Guidelines released by the American Society of Clinical Oncology (ASCO) state that ALND can be avoided in women undergoing breast conserving surgery if sentinel lymph nodes are negative for carcinoma, contain only isolated tumor cells (up to 200 tumor cells and or < 0.2 mm deposit), or contain a micrometastasis (tumor deposit of ≥ 0.2 to < 2 mm) [2]. Although intraoperative diagnoses of macrometastatic
⁎
involvement may decrease the number of additional, separate operative procedures, there is the potential for diagnostic error. Given the recent guideline changes for surgical management of the axilla, pathology laboratories need to adapt their practices to meet the changing needs of patients undergoing intraoperative axillary lymph node evaluation. There are multiple potential methods of intraoperative lymph node evaluation, including gross evaluation, imprint cytology (touch preparation), and frozen section. Intraoperative lymph node evaluation by imprint cytology maintains cytologic detail, preserves all diagnostic tissue for permanent section analysis, and allows faster intraoperative preparation of slides while showing comparable sensitivity to frozen section analysis [3]. Imprint cytology has been shown to have high specificity in intraoperative margin assessment of breast tissue and in identifying axillary lymph node metastases [4,5]. However, the impact of cytologic assessment of sentinel lymph nodes on further axillary surgical management has not been fully explored in the context of changing surgical management of the axilla. To evaluate the performance characteristics of imprint cytology, we performed a retrospective review of patients who underwent
Corresponding author at: Vanderbilt University Medical Center, 1161 21st Avenue South, MCN CC3322, Nashville, TN 37232, United States. E-mail address: [email protected] (M.L. Compton).
https://doi.org/10.1016/j.anndiagpath.2017.12.003
1092-9134/ © 2017 Elsevier Inc. All rights reserved.
Annals of Diagnostic Pathology 33 (2018) 45–50
M.L. Compton et al.
intraoperative sentinel lymph node evaluation by imprint cytology at the time of primary breast cancer excision at a single tertiary academic medical center. We then introduced a modified intraoperative protocol incorporating both gross assessment and imprint cytology to analyze the effect on distinguishing micro- from macrometastases. Sensitivity and specificity, as well as clinical impact of intraoperative lymph node evaluation by imprint cytology were determined using the modified protocol and compared to the standard protocol.
Table 1 Patient demographics and pathologic findings.
Age (years) Range Mean Surgery performed Mastectomy Partial/excision Tumor type DCISa Invasive carcinoma Completion dissection Same-day Second surgery Number of lymph nodes examined Positive lymph nodes Macrometastatic Micrometastatic Isolated tumor cells
2. Materials and methods Following Vanderbilt University Institutional Review Board approval, a retrospective review of the surgical pathology database (Cerner Copath, Kansas City, MO) was performed to identify consecutive surgical cases for excision of breast carcinoma with sentinel lymph node evaluation over a 24 month period. Intraoperative assessment for this cohort was performed by sectioning all submitted lymph nodes at 2 mm intervals and performing imprint cytology on one side of each 2 mm slice. One touchprep was performed on each slice, and the resulting slides were stained with hematoxylin and eosin (H&E). Three diagnostic categories were used under this standard protocol: 1) “negative,” 2) “atypical,” or 3) “positive.” All intraoperative cases were interpreted by a surgical pathology fellow with confirmation by an attending surgical pathologist as needed. A modified protocol was introduced in July 2014 under which all sentinel lymph nodes were sectioned at 2 mm intervals, grossly examined for the presence of suspicious lesions by visual inspection and palpation, and imprint cytology was performed. This protocol contained four diagnostic categories: 1) “negative for carcinoma” if no tumor cells were present on touch prep regardless of gross findings, 2) “atypical cells, not diagnostic of carcinoma” if atypical cells were seen on touch prep regardless of gross findings, 3) “tumor cells present but no definitive macrometastatic disease” if tumor cells were present on touch prep without gross features of macrometastasis, and 4) “tumor cells consistent with macrometastatic carcinoma” if tumor cells were identified on touch prep and gross features of macrometastasis were present. A database search was performed to identify surgical cases for excision of breast carcinoma with sentinel lymph node evaluation using this modified protocol over a 12 month period. Clinical and pathologic data including patient age, tumor size, histologic type, lymph node status, type of surgery performed and decision to perform same-day ALND were collected for both cohorts. Patients who received neoadjuvant therapy or had a non-epithelial breast malignancy were excluded from the analysis. Resource utilization was analyzed using average intraoperative turn-around time.
P values
Standard protocol (n = 213)
Modified Protocol (n = 74)
27–80 57.0
34–81 56.6
171 (80.3%) 42 (19.7%)
61 (82.4%) 13 (17.6%)
48 (22.5%) 165 (77.5%)
13 (17.6%) 61 (82.4%)
15 (7%) 6 (2.8%) 516 (2.4 per patient)
4 (5.4%) 203 (2.7 per patient)
24 (11.3%) 10 (4.7%) 0
8 (3.9%) 4 (2.0%) 2 (1.0%)
0.78
0.74
0.41
0.79 0.70
0.84 0.99 0.08
Abbreviations. a DCIS, ductal carcinoma in situ. Table 2 Comparison of results from intraoperative imprint cytology analysis vs. permanent section. Intraoperative result
Standard protocol Negative (n = 481) Atypical (n = 8) Positive (n = 27) Modified protocol Negative (n = 193) Atypical cells (n = 3) Tumor cells present, not grossly positive (n = 4) Tumor cell present, grossly positive (n = 3)
Permanent result Negative
ITCsa
Micrometastasis
Macrometastasis
460 6 0
0 0 0
5 2 3
16 0 24
187 2 0
1 0 1
3 0 1
2 1 2
0
0
0
3
Abbreviations. a ITCs, Isolated tumor cells.
shown in Fig. 1. As expected, incorporating gross examination in the modified protocol resulted in reduced sensitivity for the identification of macrometastases, but achieved the desired 100% specificity and positive predictive value (Table 4). Although no negative lymph nodes were falsely called positive using the standard protocol, touch imprint did not distinguish between micro- and macrometastatic disease. Under the standard protocol, 15 patients underwent same-day ALND, of which, 11 had macrometastatic disease and 4 had only isolated micrometastatic disease. With the modified protocol, all patients undergoing same-day ALND (n = 4), had macrometastases. A second surgery was required for completion ALND in 3 patients/year under both the standard and modified protocols. Of the patients who underwent same-day ALND under the standard system, 6 patients had additional positive non-sentinel lymph nodes. Of patients who underwent completion ALND at a later date in the standard protocol cohort, only 1 positive non-sentinel lymph node was identified. Under the modified system, of the 4 patients who underwent same-day ALND, only 1 patient had additional positive non-sentinel lymph nodes. Among the patients who underwent later completion
3. Results Sentinel lymph nodes from 602 patients were excised during the entire 3 year study period, of which 287 patients received intraoperative assessment of the nodes. Five hundred sixteen lymph nodes were analyzed in 213 patients under the standard protocol, and 203 lymph nodes were analyzed in 74 patients using the modified protocol. Patient characteristics for both cohorts were not statistically different (Table 1). Although lymph nodes were evaluated more frequently in patients undergoing mastectomy, there was no statistical difference in the rate of lymph node micrometastases (p = 0.99) or macrometastasis (p = 0.84) between the two groups. Though the absolute number of patients with macrometastatic disease were low (12 patients/year under the standard protocol and 8 patients/year under the modified protocol), macrometastases were missed by both methods (Table 2). Sources of error included lack of transfer of malignant cells in 8 cases and misinterpretation of tumor cells in 6 cases, 4 of which were invasive lobular carcinomas or carcinoma with lobular features (Table 3). Examples of intraoperative and permanent section correlates for different tumor morphologies are 46
1/3 2/2 1/3 1/3
9 10 11 12
47 6, 1.6
2/2 1/4
5
Abbreviations. a FFPE, Formalin Fixed, Paraffin Embedded. b IMC NST, invasive mammary carcinoma, no special type. c ENE, extranodal extension. d ILC, invasive lobular carcinoma.
5
4 7
1/1 1/1
Atypical, not diagnostic of malignancy
3.5
1/4
4 3.5 and 2.5 9 3 with ENE
5 4 3 9 with ENEc 7 3 8 with ENE 4 with ENE
Metastasis size (mm)
Modified protocol 1 Tumor cells present, but no definitive macrometastic disease 2 Benign 3 Tumor cells present, but no definitive macrometastic disease 4 Benign
0/3 0/2 0/3 0/3
2/3 1/2 1/2 1/2 1/2 1/4 1/2 1/3
FFPEa result
Standard protocol 1 1/3 2 0/2 3 0/2 4 0/2 5 0/2 6 0/3 7 0/2 8 0/3
Intraoperative result
Table 3 Discordant cases of macrometastatic nodal disease.
IMC NST with tubular features
IMC NST with mucinous features
IMC NST IMC NST with lobular features
IMC NST
IMC NSTb IMC NST IMC NST ILCd IMC NST IMC NST ILC IMC NST with lobular and tubular features IMC NST ILC IMC NST with lobular features IMC NST with lobular features
Histologic type
Y
Y
N N
N
N Y Y N
N Y N Y Y Y N N
Second surgery
1/27
2/20
1/10 1/1
1/7
0/5 2/17 1/24 1/3
2/26 1/7 1/2 2/32 0/16 1/27 4/5 1/15
Final node status
Tumor present on touch prep, but misinterpreted. Did not grossly identify macrometastasis. Did not grossly identify macrometastasis.
No transfer of tumor cells. Did not grossly identify macrometastasis.
Did not grossly identify macrometastasis.
Tumor present on touch prep, but misinterpreted. Tumor present on touch prep, but misinterpreted. No transfer of tumor cells. No transfer of tumor cells.
No transfer of tumor cells. No transfer of tumor cells. No transfer of tumor cells. Tumor present on touch prep, but misinterpreted. No transfer of tumor cells. No transfer of tumor cells. Tumor present on touch prep, but misinterpreted. Tumor present on touch prep, but misinterpreted.
Reason for discrepancy
M.L. Compton et al.
Annals of Diagnostic Pathology 33 (2018) 45–50
Annals of Diagnostic Pathology 33 (2018) 45–50
M.L. Compton et al.
a
b
c
d
e
f
Fig. 1. Intraoperative touch preps (a, c, e) with correlating paraffin embedded histology demonstrating micrometastatic invasive mammary carcinoma (b), macrometastatic invasive mammary carcinoma, no special type (d), and macrometastatic invasive lobular carcinoma (f) on final permanent sections.
second surgeries for completion ALND. A small but clinically significant number of women received same-day ALND based on results of intraoperative analysis, suggesting a continued role for intraoperative examination of sentinel lymph nodes. The sensitivities for detecting macrometastases achieved with both the standard (60%) and modified techniques (38%) fall within the sensitivity range for imprint cytology previously reported by other authors, which varied from 23.8% to 81.3% [3,6-11]. In prior studies, decreased sensitivity was noted with micrometastatic disease and neoadjuvant chemothetherapy [3,7,10]. Our analysis specifically excluded patients who had received neoadjuvant chemotherapy. We found that despite low sensitivity, micrometastases and isolated tumor cells were detected by imprint cytology, which could lead to unnecessary axillary dissection. In addition, there were several cases of macrometastatic disease, including lymph nodes with extranodal extension of tumor, in which there was no transfer of tumor cells on imprint cytology. It is also notable that clinical staging techniques missed large metastatic foci in several women, including tumors up to 9 mm with extranodal extension. Another potential diagnostic pitfall was invasive lobular carcinoma, as we had several cases in which the discohesive nature of the tumor cells led to misinterpretation on intraoperative evaluation, despite adequate transfer of tumor cells to the slide. Reported specificities of imprint cytology in prior studies have been high, ranging from to 92.9% to 100% [7,11]. However, these prior analyses did not assess specificity and clinical utility in the context of distinguishing micro- and macrometastatic disease. Specificity, which was calculated based on correct identification of macrometastatic
Table 4 Performance characteristics of imprint cytology by standard and modified protocols.
Sensitivity Specificity Positive predictive value Negative predictive value
Standard protocol, n = 516 (95% confidence interval)
Modified protocol, n = 203 (95% confidence interval)
60% 99% 88% 96%
38% (10–74) 100% (98–100) 100% (31–100) 96% (94–99)
(43–75) (98–100) (70–97) (95–98)
ALND in the modified cohort, no patients had additional micro- or macrometastases, and one patient had isolated tumor cells. The average turn-around time for performance of touch imprint and microscopic evaluation was 18 min per sentinel lymph node over the study period; meaning, over 4000 min of laboratory time per year was devoted to analysis of sentinel lymph nodes for breast carcinoma at our institution. 4. Discussion Imprint cytology alone could not reliably distinguish between micro- and macrometastatic disease, which resulted in same-day ALND in women with isolated micrometastases. Introduction of a modified intraoperative protocol incorporating gross examination and imprint cytology achieved 100% specificity and positive predictive value for macrometastasis at the cost of decreased sensitivity. Despite the reduced sensitivity of the method, there was no increase in the rate of 48
Annals of Diagnostic Pathology 33 (2018) 45–50
M.L. Compton et al.
rate of axillary recurrence in patients who underwent mastectomy with radiation therapy vs. modified radical mastectomy [17,18]. Currently, there is not clear evidence of what the optimal intraoperative management of this patient population would be to achieve optimal balance between avoiding unnecessary axillary dissections, minimizing risk of recurrence, and achieving cosmetically acceptable reconstruction.
disease in our study, was improved by incorporating gross evaluation. We demonstrate that positive predictive value of imprint cytology alone for macrometastases is relatively low (88%), and may lead to an otherwise unnecessary ALND for micrometastatic disease. The modified system incorporating gross analysis with imprint cytology demonstrated a clear advantage in distinguishing micro- and macro-metastatic disease. All women who underwent same-day ALND had macrometastatic disease confirmed on permanent section. However, although there was no increase in the rate of second surgeries for completion ALND, the absolute number of patients with macrometastatic disease was small in the current analysis, so the true clinical impact of the decreased sensitivity is not clear at this time. False negative intraoperative evaluation has implications for surgical management of the patient. In addition to informing the decision of whether or not to perform same-day ALND, the results of the intraoperative assessment impacts reconstruction options [12]. Post mastectomy radiation can have a significant affect on reconstructive decision making. Radiation of autologous flaps should be avoided. With increasing recommendations for axillary radiation in patients with 1–3 positive nodes, the status of the sentinel nodes is critical to reconstructive planning, driving some centers to perform a separate sentinel node biopsy prior to the index procedure. Additionally, ALND after autologous flap reconstruction can compromise the vascular pedicle [13,14]. ALND has also been found to be associated with a higher rate of implant loss and infection in patients who underwent same-day tissue expander implantation with subsequent radiation [15]. Surgeons must counsel patients that the intraoperative assessment is not the definitive pathologic diagnosis. With regard to resource utilization, despite the finding that intraoperative lymph node assessment consumes substantial staffing time in the surgical pathology area, it does not necessarily extend the operative time for the patient, as other aspects of the surgery may be performed while awaiting the pathology results [10]. Our retrospective study was not able to adequately assess the effect of intraoperative evaluation on operative times since multiple factors influence operative time. Similarly, although intraoperative cytologic assessment is associated with a relatively substantial cost of approximately $320 per specimen based on Medicare reimbursement rates for professional and technical fees, any potential cost savings could be negated if the number of second surgeries increases. Since imprint cytology of grossly negative lymph nodes did not provide additional information for clinical management of the patient under our modified protocol, surgical pathologists and surgeons may consider performing cytologic assessment of only those lymph nodes that are grossly suspicious for macrometastases. This would conserve resources and decrease billing, as gross assessment of a lymph node results in a bill of only $16 per specimen. Given our laboratory's volume, performing imprint cytology on only grossly positive lymph nodes could save an estimated $150,000 per year. Due to changes in clinical guidelines regarding axillary management, it is important to re-evaluate the performance of intraoperative nodal evaluation in this context. The importance of intraoperative assessment has shifted to specific populations. As a result of the practice changing adoption of Z0011 criteria, the majority of women who received intraoperative assessment of lymph node status at our institution underwent mastectomy. The evidence regarding axillary management in women undergoing mastectomy who are found to have isolated micrometastases is evolving. ASCO and National Comprehensive Cancer Network (NCCN) recommend ALND dissection in this population; however, the evidence from the IBCSG 23-01 trial suggest equivalent disease free survival in women who forgo ALND as compared to women who receive ALND when there is micrometastatic nodal disease alone [16]. The AMAROS trial suggested non-inferiority for avoiding ALND in patients with a positive sentinel lymph node, but the study was underpowered due to low rates of axillary recurrence in both groups. In contrast, the NSABP B-04 trial found a slightly higher
5. Conclusions/concluding remarks Imprint cytology is a technique with high specificity for identification of tumor cells; however, it does not distinguish between isolated tumor cells, micro-, and macrometastatic disease, and may lead to ALND for isolated micrometastases. With the surgical management of axillary lymph nodes in breast carcinoma becoming progressively more conservative, determination of tumor deposit sizes in sentinel lymph nodes has become increasingly important. The method of intraoperative evaluation must incorporate this information to provide surgeons the level of information needed to make next step surgical decisions. Our study found that incorporation of gross examination of lymph nodes yields a technique with high specificity and positive predictive value for identifying macrometastases, but demonstrates a resultant decrease in sensitivity. Pathologists and surgical oncologists should be aware of the limitations of intraoperative evaluation techniques, as this may have a significant impact on surgical planning. Conflicts of interest None. Funding This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sections. References [1] Giuliano AE, Hunt KK, Ballman KV, et al. Axillary dissection vs no axillary dissection in women with invasive breast cancer and sentinel node metastasis a randomized clinical trial. JAMA 2011;305(6):569–75. [2] Lyman GH, Temin S, Edge SB, Newman LA, et al. Sentinel lymph node biopsy for patients with early-stage breast cancer: American Society of Clinical Oncology clinical practice guideline update. J Clin Oncol May 1 2014;32(13):1365–83. [3] Arlicot C, Louarn A, Arbion F, et al. Evaluation of the two intraoperative examination methods for sentinel lymph node assessment: a multicentric and retrospective study on more than 2,000 nodes. Anticancer Res 2013;33(3):1045–52. [4] St John ER, Al-Khudairi R, Ashrafian H, et al. Diagnostic accuracy of intraoperative techniques for margin assessment in breast cancer surgery: a meta-analysis. Ann Surg 2017 Feb;265(2):300–10. [5] Mannell A, Wium B, Thatcher C. Intraoperative examination of sentinel lymph nodes using scrape cytology. S Afr J Surg 2014;52(3):75–8. [6] Mullenix PS, Carter PL, Martin MJ, et al. Predictive value of intraoperative touch preparation analysis of sentinel lymph nodes for axillary metastasis in breast cancer. Am J Surg 2003;185(5):420–4. [7] Elliott RM, Shenk RR, Thompson CL, Gilmore HL. Touch preparations for the intraoperative evaluation of sentinel lymph nodes after neoadjuvant therapy have high false-negative rates in patients with breast cancer. Arch Pathol Lab Med 2014 Jun;138(6):814–8. [8] Chen JJ, Chen JY, Yang BL, et al. Comparison of molecular analysis and touch imprint cytology for the intraoperative evaluation of sentinel lymph nodes in primary breast cancer: results of the China Breast Cancer Clinical Study Group (CBCSG) 001c trial. Eur J Surg Oncol 2013;39:442–9. [9] Lumachi F, Marino F, Zanella S, Chiara GB, Basso SMM. Touch imprint cytology and frozen-section analysis for intraoperative evaluation of sentinel nodes in early breast cancer. Anticancer Res 2012;32:3523–6. [10] Richards ADM, Lakhani SR, James DT, Ung OA. Intraoperative imprint cytology for breast cancer sentinel nodes: is it worth it? ANZ J Surg 2013;83:39–544. [11] Wang YS, Ou-Yang T, Wu J, et al. Comparative study of one-step nucleic acid amplification assay, frozen section, and touch imprint cytology for intraoperative assessment of breast sentinel lymph node in Chinese patients. Cancer Sci 2012;103(11):1989–93. [12] Kronowitz SJ, Kuerer HM. Advances and surgical decision-making for breast reconstruction. Cancer 2006;107(5):893–907. [13] Mannu GS, Navi A, Morgan A, et al. Sentinel lymph node biopsy before mastectomy and immediate breast reconstruction may predict post-mastectomy radiotherapy, reduce delayed complications and improve the choice of reconstruction. Int J Surg
49
Annals of Diagnostic Pathology 33 (2018) 45–50
M.L. Compton et al.
node micrometastases. Lancet Oncol 2013;14(4):297–305. [17] Donker M, van Tienhoven G, Straver ME, et al. Radiotherapy or surgery of the axilla after a positive sentinel node in breast cancer (EORTC 10981-22023 AMAROS): a randomised, multicentre, open-label, phase 3 non-inferiority trial. Lancet Oncol 2014 Nov;15(12):1303–10. [18] Fisher B, Redmond C, Fisher ER, et al. 10-year results of a randomized clinical trial comparing radical mastectomy and total mastectomy with or without radiation. N Engl J Med 1985;312:674–81.
2012;10:259–64. [14] Mannu GS, Navi A, Hussien M. Sentinel lymph node biopsy before mastectomy and immediate breast reconstruction does not significantly delay surgery in early breast cancer. ANZ J Surg 2015;85:438–43. [15] Wang F, Peled A, Chin R, et al. The impact of radiation therapy, lymph node dissection, and hormonal therapy on outcomes of tissue expander–implant exchange in prosthetic breast reconstruction. Plast Reconstr Surg 2016;137(1):1–9. [16] Galimberti V, Cole BF, Zurrida S, et al. IBCSG 23-01 randomised controlled trial comparing axillary dissection versus no axillary dissection in patients with sentinel
50
Contents lists available at ScienceDirect
Annals of Diagnostic Pathology journal homepage: www.elsevier.com/locate/anndiagpath
Intraoperative sentinel lymph node evaluation: Optimizing surgical pathology practices in an era of changing clinical management
T
⁎
Margaret L. Comptona, , Raeshell S. Sweetingb, Emily S. Reisenbichlera,c a
Department of Pathology, Microbiology and Immunology, Vanderbilt University Medical Center, 1161 21st Avenue South CC-3322 Medical Center North, Nashville, TN 37232-2561, United States b Department of Surgery, Vanderbilt University Medical Center, 1161 21st Avenue South CC-3322 Medical Center North, Nashville, TN 37232-2561, United States c Yale University, Department of Pathology, 310 Cedar Street, PO Box 208023, New Haven, CT 06520-8023, United States
A R T I C L E I N F O
A B S T R A C T
Keywords: Breast Axilla Sentinel lymph node Intraoperative evaluation Surgical pathology
Axillary lymph node status is an independent prognostic indicator in breast cancer. Intraoperative identification of metastatic carcinoma in sentinel lymph nodes may allow for concurrent axillary lymph node dissection at the time of primary tumor excision. A retrospective review of patients undergoing primary breast cancer excision with sentinel lymph node sampling was performed. Sensitivity and specificity of imprint cytology (touch prep) with and without the incorporation of gross evaluation was determined using permanent section results as the gold standard. Five hundred sixteen lymph nodes were analyzed by imprint cytology in 213 patients, and 203 lymph nodes were analyzed in 74 patients incorporating gross examination. Sensitivity and specificity for the detection of macrometastases by touch prep alone were 60% and 99% respectively with 4 patients undergoing same-day axillary dissection for only micrometastatic disease. False negative causes included lack of transfer of malignant cells in 8 cases and misinterpretation of tumor cells in 6 cases. Incorporating gross examination in the modified protocol resulted in reduced sensitivity of 38%, but achieved the desired 100% specificity and positive predictive value. Imprint cytology alone did not reliably distinguish between micro- and macrometastatic disease. Gross assessment combined with imprint cytology allows for improved assessment of volume of axillary disease, but is an insensitive technique.
1. Introduction Axillary lymph node status is an independent prognostic indicator in breast cancer. In patients with a clinically negative axilla, intraoperative identification of macrometastatic carcinoma (> 2 mm tumor deposit) in sentinel lymph nodes allows for concurrent axillary lymph node dissection (ALND) and complete pathologic staging at the time of primary tumor excision. However, ALND may be associated with morbidity including pain, nerve damage, loss of mobility, lymphedema and, per the American College of Surgeons Oncology Group (ACOSOG) Z0011 clinical trial and the International Breast Cancer Study Group 23-01 clinical trial, has not been shown to improve overall survival in selected groups of women with isolated metastatic disease [1]. Guidelines released by the American Society of Clinical Oncology (ASCO) state that ALND can be avoided in women undergoing breast conserving surgery if sentinel lymph nodes are negative for carcinoma, contain only isolated tumor cells (up to 200 tumor cells and or < 0.2 mm deposit), or contain a micrometastasis (tumor deposit of ≥ 0.2 to < 2 mm) [2]. Although intraoperative diagnoses of macrometastatic
⁎
involvement may decrease the number of additional, separate operative procedures, there is the potential for diagnostic error. Given the recent guideline changes for surgical management of the axilla, pathology laboratories need to adapt their practices to meet the changing needs of patients undergoing intraoperative axillary lymph node evaluation. There are multiple potential methods of intraoperative lymph node evaluation, including gross evaluation, imprint cytology (touch preparation), and frozen section. Intraoperative lymph node evaluation by imprint cytology maintains cytologic detail, preserves all diagnostic tissue for permanent section analysis, and allows faster intraoperative preparation of slides while showing comparable sensitivity to frozen section analysis [3]. Imprint cytology has been shown to have high specificity in intraoperative margin assessment of breast tissue and in identifying axillary lymph node metastases [4,5]. However, the impact of cytologic assessment of sentinel lymph nodes on further axillary surgical management has not been fully explored in the context of changing surgical management of the axilla. To evaluate the performance characteristics of imprint cytology, we performed a retrospective review of patients who underwent
Corresponding author at: Vanderbilt University Medical Center, 1161 21st Avenue South, MCN CC3322, Nashville, TN 37232, United States. E-mail address: [email protected] (M.L. Compton).
https://doi.org/10.1016/j.anndiagpath.2017.12.003
1092-9134/ © 2017 Elsevier Inc. All rights reserved.
Annals of Diagnostic Pathology 33 (2018) 45–50
M.L. Compton et al.
intraoperative sentinel lymph node evaluation by imprint cytology at the time of primary breast cancer excision at a single tertiary academic medical center. We then introduced a modified intraoperative protocol incorporating both gross assessment and imprint cytology to analyze the effect on distinguishing micro- from macrometastases. Sensitivity and specificity, as well as clinical impact of intraoperative lymph node evaluation by imprint cytology were determined using the modified protocol and compared to the standard protocol.
Table 1 Patient demographics and pathologic findings.
Age (years) Range Mean Surgery performed Mastectomy Partial/excision Tumor type DCISa Invasive carcinoma Completion dissection Same-day Second surgery Number of lymph nodes examined Positive lymph nodes Macrometastatic Micrometastatic Isolated tumor cells
2. Materials and methods Following Vanderbilt University Institutional Review Board approval, a retrospective review of the surgical pathology database (Cerner Copath, Kansas City, MO) was performed to identify consecutive surgical cases for excision of breast carcinoma with sentinel lymph node evaluation over a 24 month period. Intraoperative assessment for this cohort was performed by sectioning all submitted lymph nodes at 2 mm intervals and performing imprint cytology on one side of each 2 mm slice. One touchprep was performed on each slice, and the resulting slides were stained with hematoxylin and eosin (H&E). Three diagnostic categories were used under this standard protocol: 1) “negative,” 2) “atypical,” or 3) “positive.” All intraoperative cases were interpreted by a surgical pathology fellow with confirmation by an attending surgical pathologist as needed. A modified protocol was introduced in July 2014 under which all sentinel lymph nodes were sectioned at 2 mm intervals, grossly examined for the presence of suspicious lesions by visual inspection and palpation, and imprint cytology was performed. This protocol contained four diagnostic categories: 1) “negative for carcinoma” if no tumor cells were present on touch prep regardless of gross findings, 2) “atypical cells, not diagnostic of carcinoma” if atypical cells were seen on touch prep regardless of gross findings, 3) “tumor cells present but no definitive macrometastatic disease” if tumor cells were present on touch prep without gross features of macrometastasis, and 4) “tumor cells consistent with macrometastatic carcinoma” if tumor cells were identified on touch prep and gross features of macrometastasis were present. A database search was performed to identify surgical cases for excision of breast carcinoma with sentinel lymph node evaluation using this modified protocol over a 12 month period. Clinical and pathologic data including patient age, tumor size, histologic type, lymph node status, type of surgery performed and decision to perform same-day ALND were collected for both cohorts. Patients who received neoadjuvant therapy or had a non-epithelial breast malignancy were excluded from the analysis. Resource utilization was analyzed using average intraoperative turn-around time.
P values
Standard protocol (n = 213)
Modified Protocol (n = 74)
27–80 57.0
34–81 56.6
171 (80.3%) 42 (19.7%)
61 (82.4%) 13 (17.6%)
48 (22.5%) 165 (77.5%)
13 (17.6%) 61 (82.4%)
15 (7%) 6 (2.8%) 516 (2.4 per patient)
4 (5.4%) 203 (2.7 per patient)
24 (11.3%) 10 (4.7%) 0
8 (3.9%) 4 (2.0%) 2 (1.0%)
0.78
0.74
0.41
0.79 0.70
0.84 0.99 0.08
Abbreviations. a DCIS, ductal carcinoma in situ. Table 2 Comparison of results from intraoperative imprint cytology analysis vs. permanent section. Intraoperative result
Standard protocol Negative (n = 481) Atypical (n = 8) Positive (n = 27) Modified protocol Negative (n = 193) Atypical cells (n = 3) Tumor cells present, not grossly positive (n = 4) Tumor cell present, grossly positive (n = 3)
Permanent result Negative
ITCsa
Micrometastasis
Macrometastasis
460 6 0
0 0 0
5 2 3
16 0 24
187 2 0
1 0 1
3 0 1
2 1 2
0
0
0
3
Abbreviations. a ITCs, Isolated tumor cells.
shown in Fig. 1. As expected, incorporating gross examination in the modified protocol resulted in reduced sensitivity for the identification of macrometastases, but achieved the desired 100% specificity and positive predictive value (Table 4). Although no negative lymph nodes were falsely called positive using the standard protocol, touch imprint did not distinguish between micro- and macrometastatic disease. Under the standard protocol, 15 patients underwent same-day ALND, of which, 11 had macrometastatic disease and 4 had only isolated micrometastatic disease. With the modified protocol, all patients undergoing same-day ALND (n = 4), had macrometastases. A second surgery was required for completion ALND in 3 patients/year under both the standard and modified protocols. Of the patients who underwent same-day ALND under the standard system, 6 patients had additional positive non-sentinel lymph nodes. Of patients who underwent completion ALND at a later date in the standard protocol cohort, only 1 positive non-sentinel lymph node was identified. Under the modified system, of the 4 patients who underwent same-day ALND, only 1 patient had additional positive non-sentinel lymph nodes. Among the patients who underwent later completion
3. Results Sentinel lymph nodes from 602 patients were excised during the entire 3 year study period, of which 287 patients received intraoperative assessment of the nodes. Five hundred sixteen lymph nodes were analyzed in 213 patients under the standard protocol, and 203 lymph nodes were analyzed in 74 patients using the modified protocol. Patient characteristics for both cohorts were not statistically different (Table 1). Although lymph nodes were evaluated more frequently in patients undergoing mastectomy, there was no statistical difference in the rate of lymph node micrometastases (p = 0.99) or macrometastasis (p = 0.84) between the two groups. Though the absolute number of patients with macrometastatic disease were low (12 patients/year under the standard protocol and 8 patients/year under the modified protocol), macrometastases were missed by both methods (Table 2). Sources of error included lack of transfer of malignant cells in 8 cases and misinterpretation of tumor cells in 6 cases, 4 of which were invasive lobular carcinomas or carcinoma with lobular features (Table 3). Examples of intraoperative and permanent section correlates for different tumor morphologies are 46
1/3 2/2 1/3 1/3
9 10 11 12
47 6, 1.6
2/2 1/4
5
Abbreviations. a FFPE, Formalin Fixed, Paraffin Embedded. b IMC NST, invasive mammary carcinoma, no special type. c ENE, extranodal extension. d ILC, invasive lobular carcinoma.
5
4 7
1/1 1/1
Atypical, not diagnostic of malignancy
3.5
1/4
4 3.5 and 2.5 9 3 with ENE
5 4 3 9 with ENEc 7 3 8 with ENE 4 with ENE
Metastasis size (mm)
Modified protocol 1 Tumor cells present, but no definitive macrometastic disease 2 Benign 3 Tumor cells present, but no definitive macrometastic disease 4 Benign
0/3 0/2 0/3 0/3
2/3 1/2 1/2 1/2 1/2 1/4 1/2 1/3
FFPEa result
Standard protocol 1 1/3 2 0/2 3 0/2 4 0/2 5 0/2 6 0/3 7 0/2 8 0/3
Intraoperative result
Table 3 Discordant cases of macrometastatic nodal disease.
IMC NST with tubular features
IMC NST with mucinous features
IMC NST IMC NST with lobular features
IMC NST
IMC NSTb IMC NST IMC NST ILCd IMC NST IMC NST ILC IMC NST with lobular and tubular features IMC NST ILC IMC NST with lobular features IMC NST with lobular features
Histologic type
Y
Y
N N
N
N Y Y N
N Y N Y Y Y N N
Second surgery
1/27
2/20
1/10 1/1
1/7
0/5 2/17 1/24 1/3
2/26 1/7 1/2 2/32 0/16 1/27 4/5 1/15
Final node status
Tumor present on touch prep, but misinterpreted. Did not grossly identify macrometastasis. Did not grossly identify macrometastasis.
No transfer of tumor cells. Did not grossly identify macrometastasis.
Did not grossly identify macrometastasis.
Tumor present on touch prep, but misinterpreted. Tumor present on touch prep, but misinterpreted. No transfer of tumor cells. No transfer of tumor cells.
No transfer of tumor cells. No transfer of tumor cells. No transfer of tumor cells. Tumor present on touch prep, but misinterpreted. No transfer of tumor cells. No transfer of tumor cells. Tumor present on touch prep, but misinterpreted. Tumor present on touch prep, but misinterpreted.
Reason for discrepancy
M.L. Compton et al.
Annals of Diagnostic Pathology 33 (2018) 45–50
Annals of Diagnostic Pathology 33 (2018) 45–50
M.L. Compton et al.
a
b
c
d
e
f
Fig. 1. Intraoperative touch preps (a, c, e) with correlating paraffin embedded histology demonstrating micrometastatic invasive mammary carcinoma (b), macrometastatic invasive mammary carcinoma, no special type (d), and macrometastatic invasive lobular carcinoma (f) on final permanent sections.
second surgeries for completion ALND. A small but clinically significant number of women received same-day ALND based on results of intraoperative analysis, suggesting a continued role for intraoperative examination of sentinel lymph nodes. The sensitivities for detecting macrometastases achieved with both the standard (60%) and modified techniques (38%) fall within the sensitivity range for imprint cytology previously reported by other authors, which varied from 23.8% to 81.3% [3,6-11]. In prior studies, decreased sensitivity was noted with micrometastatic disease and neoadjuvant chemothetherapy [3,7,10]. Our analysis specifically excluded patients who had received neoadjuvant chemotherapy. We found that despite low sensitivity, micrometastases and isolated tumor cells were detected by imprint cytology, which could lead to unnecessary axillary dissection. In addition, there were several cases of macrometastatic disease, including lymph nodes with extranodal extension of tumor, in which there was no transfer of tumor cells on imprint cytology. It is also notable that clinical staging techniques missed large metastatic foci in several women, including tumors up to 9 mm with extranodal extension. Another potential diagnostic pitfall was invasive lobular carcinoma, as we had several cases in which the discohesive nature of the tumor cells led to misinterpretation on intraoperative evaluation, despite adequate transfer of tumor cells to the slide. Reported specificities of imprint cytology in prior studies have been high, ranging from to 92.9% to 100% [7,11]. However, these prior analyses did not assess specificity and clinical utility in the context of distinguishing micro- and macrometastatic disease. Specificity, which was calculated based on correct identification of macrometastatic
Table 4 Performance characteristics of imprint cytology by standard and modified protocols.
Sensitivity Specificity Positive predictive value Negative predictive value
Standard protocol, n = 516 (95% confidence interval)
Modified protocol, n = 203 (95% confidence interval)
60% 99% 88% 96%
38% (10–74) 100% (98–100) 100% (31–100) 96% (94–99)
(43–75) (98–100) (70–97) (95–98)
ALND in the modified cohort, no patients had additional micro- or macrometastases, and one patient had isolated tumor cells. The average turn-around time for performance of touch imprint and microscopic evaluation was 18 min per sentinel lymph node over the study period; meaning, over 4000 min of laboratory time per year was devoted to analysis of sentinel lymph nodes for breast carcinoma at our institution. 4. Discussion Imprint cytology alone could not reliably distinguish between micro- and macrometastatic disease, which resulted in same-day ALND in women with isolated micrometastases. Introduction of a modified intraoperative protocol incorporating gross examination and imprint cytology achieved 100% specificity and positive predictive value for macrometastasis at the cost of decreased sensitivity. Despite the reduced sensitivity of the method, there was no increase in the rate of 48
Annals of Diagnostic Pathology 33 (2018) 45–50
M.L. Compton et al.
rate of axillary recurrence in patients who underwent mastectomy with radiation therapy vs. modified radical mastectomy [17,18]. Currently, there is not clear evidence of what the optimal intraoperative management of this patient population would be to achieve optimal balance between avoiding unnecessary axillary dissections, minimizing risk of recurrence, and achieving cosmetically acceptable reconstruction.
disease in our study, was improved by incorporating gross evaluation. We demonstrate that positive predictive value of imprint cytology alone for macrometastases is relatively low (88%), and may lead to an otherwise unnecessary ALND for micrometastatic disease. The modified system incorporating gross analysis with imprint cytology demonstrated a clear advantage in distinguishing micro- and macro-metastatic disease. All women who underwent same-day ALND had macrometastatic disease confirmed on permanent section. However, although there was no increase in the rate of second surgeries for completion ALND, the absolute number of patients with macrometastatic disease was small in the current analysis, so the true clinical impact of the decreased sensitivity is not clear at this time. False negative intraoperative evaluation has implications for surgical management of the patient. In addition to informing the decision of whether or not to perform same-day ALND, the results of the intraoperative assessment impacts reconstruction options [12]. Post mastectomy radiation can have a significant affect on reconstructive decision making. Radiation of autologous flaps should be avoided. With increasing recommendations for axillary radiation in patients with 1–3 positive nodes, the status of the sentinel nodes is critical to reconstructive planning, driving some centers to perform a separate sentinel node biopsy prior to the index procedure. Additionally, ALND after autologous flap reconstruction can compromise the vascular pedicle [13,14]. ALND has also been found to be associated with a higher rate of implant loss and infection in patients who underwent same-day tissue expander implantation with subsequent radiation [15]. Surgeons must counsel patients that the intraoperative assessment is not the definitive pathologic diagnosis. With regard to resource utilization, despite the finding that intraoperative lymph node assessment consumes substantial staffing time in the surgical pathology area, it does not necessarily extend the operative time for the patient, as other aspects of the surgery may be performed while awaiting the pathology results [10]. Our retrospective study was not able to adequately assess the effect of intraoperative evaluation on operative times since multiple factors influence operative time. Similarly, although intraoperative cytologic assessment is associated with a relatively substantial cost of approximately $320 per specimen based on Medicare reimbursement rates for professional and technical fees, any potential cost savings could be negated if the number of second surgeries increases. Since imprint cytology of grossly negative lymph nodes did not provide additional information for clinical management of the patient under our modified protocol, surgical pathologists and surgeons may consider performing cytologic assessment of only those lymph nodes that are grossly suspicious for macrometastases. This would conserve resources and decrease billing, as gross assessment of a lymph node results in a bill of only $16 per specimen. Given our laboratory's volume, performing imprint cytology on only grossly positive lymph nodes could save an estimated $150,000 per year. Due to changes in clinical guidelines regarding axillary management, it is important to re-evaluate the performance of intraoperative nodal evaluation in this context. The importance of intraoperative assessment has shifted to specific populations. As a result of the practice changing adoption of Z0011 criteria, the majority of women who received intraoperative assessment of lymph node status at our institution underwent mastectomy. The evidence regarding axillary management in women undergoing mastectomy who are found to have isolated micrometastases is evolving. ASCO and National Comprehensive Cancer Network (NCCN) recommend ALND dissection in this population; however, the evidence from the IBCSG 23-01 trial suggest equivalent disease free survival in women who forgo ALND as compared to women who receive ALND when there is micrometastatic nodal disease alone [16]. The AMAROS trial suggested non-inferiority for avoiding ALND in patients with a positive sentinel lymph node, but the study was underpowered due to low rates of axillary recurrence in both groups. In contrast, the NSABP B-04 trial found a slightly higher
5. Conclusions/concluding remarks Imprint cytology is a technique with high specificity for identification of tumor cells; however, it does not distinguish between isolated tumor cells, micro-, and macrometastatic disease, and may lead to ALND for isolated micrometastases. With the surgical management of axillary lymph nodes in breast carcinoma becoming progressively more conservative, determination of tumor deposit sizes in sentinel lymph nodes has become increasingly important. The method of intraoperative evaluation must incorporate this information to provide surgeons the level of information needed to make next step surgical decisions. Our study found that incorporation of gross examination of lymph nodes yields a technique with high specificity and positive predictive value for identifying macrometastases, but demonstrates a resultant decrease in sensitivity. Pathologists and surgical oncologists should be aware of the limitations of intraoperative evaluation techniques, as this may have a significant impact on surgical planning. Conflicts of interest None. Funding This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sections. References [1] Giuliano AE, Hunt KK, Ballman KV, et al. Axillary dissection vs no axillary dissection in women with invasive breast cancer and sentinel node metastasis a randomized clinical trial. JAMA 2011;305(6):569–75. [2] Lyman GH, Temin S, Edge SB, Newman LA, et al. Sentinel lymph node biopsy for patients with early-stage breast cancer: American Society of Clinical Oncology clinical practice guideline update. J Clin Oncol May 1 2014;32(13):1365–83. [3] Arlicot C, Louarn A, Arbion F, et al. Evaluation of the two intraoperative examination methods for sentinel lymph node assessment: a multicentric and retrospective study on more than 2,000 nodes. Anticancer Res 2013;33(3):1045–52. [4] St John ER, Al-Khudairi R, Ashrafian H, et al. Diagnostic accuracy of intraoperative techniques for margin assessment in breast cancer surgery: a meta-analysis. Ann Surg 2017 Feb;265(2):300–10. [5] Mannell A, Wium B, Thatcher C. Intraoperative examination of sentinel lymph nodes using scrape cytology. S Afr J Surg 2014;52(3):75–8. [6] Mullenix PS, Carter PL, Martin MJ, et al. Predictive value of intraoperative touch preparation analysis of sentinel lymph nodes for axillary metastasis in breast cancer. Am J Surg 2003;185(5):420–4. [7] Elliott RM, Shenk RR, Thompson CL, Gilmore HL. Touch preparations for the intraoperative evaluation of sentinel lymph nodes after neoadjuvant therapy have high false-negative rates in patients with breast cancer. Arch Pathol Lab Med 2014 Jun;138(6):814–8. [8] Chen JJ, Chen JY, Yang BL, et al. Comparison of molecular analysis and touch imprint cytology for the intraoperative evaluation of sentinel lymph nodes in primary breast cancer: results of the China Breast Cancer Clinical Study Group (CBCSG) 001c trial. Eur J Surg Oncol 2013;39:442–9. [9] Lumachi F, Marino F, Zanella S, Chiara GB, Basso SMM. Touch imprint cytology and frozen-section analysis for intraoperative evaluation of sentinel nodes in early breast cancer. Anticancer Res 2012;32:3523–6. [10] Richards ADM, Lakhani SR, James DT, Ung OA. Intraoperative imprint cytology for breast cancer sentinel nodes: is it worth it? ANZ J Surg 2013;83:39–544. [11] Wang YS, Ou-Yang T, Wu J, et al. Comparative study of one-step nucleic acid amplification assay, frozen section, and touch imprint cytology for intraoperative assessment of breast sentinel lymph node in Chinese patients. Cancer Sci 2012;103(11):1989–93. [12] Kronowitz SJ, Kuerer HM. Advances and surgical decision-making for breast reconstruction. Cancer 2006;107(5):893–907. [13] Mannu GS, Navi A, Morgan A, et al. Sentinel lymph node biopsy before mastectomy and immediate breast reconstruction may predict post-mastectomy radiotherapy, reduce delayed complications and improve the choice of reconstruction. Int J Surg
49
Annals of Diagnostic Pathology 33 (2018) 45–50
M.L. Compton et al.
node micrometastases. Lancet Oncol 2013;14(4):297–305. [17] Donker M, van Tienhoven G, Straver ME, et al. Radiotherapy or surgery of the axilla after a positive sentinel node in breast cancer (EORTC 10981-22023 AMAROS): a randomised, multicentre, open-label, phase 3 non-inferiority trial. Lancet Oncol 2014 Nov;15(12):1303–10. [18] Fisher B, Redmond C, Fisher ER, et al. 10-year results of a randomized clinical trial comparing radical mastectomy and total mastectomy with or without radiation. N Engl J Med 1985;312:674–81.
2012;10:259–64. [14] Mannu GS, Navi A, Hussien M. Sentinel lymph node biopsy before mastectomy and immediate breast reconstruction does not significantly delay surgery in early breast cancer. ANZ J Surg 2015;85:438–43. [15] Wang F, Peled A, Chin R, et al. The impact of radiation therapy, lymph node dissection, and hormonal therapy on outcomes of tissue expander–implant exchange in prosthetic breast reconstruction. Plast Reconstr Surg 2016;137(1):1–9. [16] Galimberti V, Cole BF, Zurrida S, et al. IBCSG 23-01 randomised controlled trial comparing axillary dissection versus no axillary dissection in patients with sentinel
50