Intraoperative molecular assay for sentinel lymph node metastases in early stage breast cancer
A comparative analysis between one-step nucleic acid amplification whole node assay and routine frozen section histology
Conventional histopathological examination is limited in measuring accurate total metastatic volume in a lymph node. Recently, a molecular-based procedure to detect lymph node metastases, one-step nucleic acid amplification (OSNA) assay, has been developed. OSNA assay can assess a whole lymph node and yields semiquantitative results. The authors compared the performance in intraoperative detection of sentinel lymph node metastases with OSNA assay using a whole lymph node versus routine frozen section (FS) histology with a 2 mm-sectioned lymph node.
Subjects comprised 531 consecutive patients diagnosed with OSNA assay and 618 consecutive patients diagnosed with FS histological examination. The authors compared the sentinel lymph node-positive rate between the OSNA and FS cohorts, and investigated characteristics of patients for whom OSNA could detect metastases but FS could not. OSNA (+) was defined as micrometastasis, and OSNA (++) and (+I) were defined as macrometastasis.
OSNA assay detected more cases of sentinel lymph node metastases than FS histology (OSNA 121 of 531, 22.8% vs FS 109 of 618, 17.6%; P = .036), particularly micrometastases (46 of 531, 8.7% vs 28 of 618, 4.5%; P = .0064). There was no difference in macrometastasis detection between OSNA and FS (75 of 531, 14.1% vs 81 of 618, 13.1%; P = .68). OSNA detected more metastases than FS in postmenopausal patients (77 of 302, 25.5% vs 43 of 351, 12.3%; P < .0001), and in tumors without fat invasion (23 of 156, 14.7% vs 6 of 151, 4.0%; P = .012) or lymphovascular invasion (67 of 395, 17.0% vs 45 of 458, 9.8%; P = .042).
Intraoperative OSNA assay detects more sentinel lymph node metastases, particularly micrometastases, than does FS histology. OSNA assay can also detect more metastases in postmenopausal patients or from less aggressive primary tumors compared with FS histology. Cancer 2011. © 2011 American Cancer Society.
Axillary lymph node status is the most powerful prognostic factor in breast cancer, and knowledge of this is essential for making decisions about adjuvant therapy. Since the late 1940s, pathologists have been aware that the more sections taken from axillary lymph nodes, the more metastases are identified.1 Since then, several studies have used serial sectioning to detect occult axillary lymph node metastases in patients previously staged as lymph node-negative by conventional histology.2 However, in daily clinical practice, examining all axillary nodes by serial sectioning is not feasible.
This view changed after introduction of the sentinel lymph node concept in the early 1990s.3, 4 To prevent false-negative diagnoses, pathologists began to concentrate on the evaluation of sentinel lymph nodes by adopting a step-sectioning procedure.5 Intensive examination of sentinel lymph nodes resulted in an increase in the detection of low-volume metastases.6
Before the introduction of sentinel lymph node biopsy, all low-volume metastases found by conventional histology were classified as lymph node positive. However, in 2002 the Cancer Staging Manual of the American Joint Committee on Cancer (AJCC), sixth edition,7 classified low-volume metastases into isolated tumor cells (ITCs) and micrometastases. ITCs were staged as pN0(i+) with deposits ≤0.2 mm, and micrometastasis was staged as pN1mi with deposits >0.2 mm to ≤2 mm. Moreover, in the seventh edition of the AJCC Staging Manual in 2010,8 T1 with lymph node spread confined to micrometastasis (pN1mi) has been down-staged from stage IIA to stage IB. Therefore, accurate evaluation of the metastatic volume in a lymph node, in terms of ITCs, micrometastasis, or macrometastasis, has been important.
Conventional histopathological examinations are limited in their ability to accurately measure total metastatic volume in a lymph node. Even if a lymph node is step-sectioned and histologically evaluated at each cut surface, the information gathered is incomplete because only a small part of the lymph node is analyzed. Although the false-negative rate and underestimation of the metastasis volume can be reduced by serial sectioning and immunohistochemistry, this imposes a heavy workload for pathologists, and it is not available as an intraoperative evaluation. Intraoperative frozen section (FS) histology has been widely used for rapid diagnosis of the sentinel lymph node status. However, the false-negative rate of FS histology, mainly resulting from failure to detect micrometastases, have been found to range from 26% to 43% compared with the final pathological results.9
The one-step nucleic acid amplification (OSNA) assay (Sysmex, Kobe, Japan) is a rapid molecular detection procedure that analyses lymph node metastases by detection and amplification of cytokeratin 19 (CK19) mRNA.10 This assay could accurately detect clinically relevant nodal metastases of >0.2 mm at rates comparable to conventional pathological examination, in at least half of the lymph nodes tested.10-13 This assay can assess a whole lymph node and yields semiquantitative results. Therefore, to evaluate the total volume of metastases, a whole lymph node should be used, leaving none of the lymph node for cytological or histological examination.
Since April 2009, we have introduced the OSNA assay as a method of intraoperative sentinel lymph node diagnosis and have used a whole lymph node in the assay. In the present retrospective cohort study, we compare performance of the intraoperative detection of sentinel lymph node metastases with OSNA assay versus routine FS histology using a 2 mm-sectioned lymph node, and reveal characteristics of OSNA assay-positive patients.
MATERIALS AND METHODS
Patients and Tumors
Subjects comprised consecutive patients with clinically and ultrasonographically lymph node-negative pT1-2 breast cancer who underwent sentinel lymph node biopsy with a radioisotope (RI) tracer between January and December 2008 and between April 2009 and March 2010 at the Cancer Institute Hospital, Tokyo, Japan. The earlier patients underwent FS histology (FS cohort), and the later patients underwent OSNA assay (OSNA cohort) for the intraoperative detection of sentinel lymph node metastasis. Exclusion criteria were as follows: 1) sentinel lymph node identification with dye alone, 2) primary tumor previously excised, 3) metastasis detected in only a nonsentinel lymph node, 4) heterochronous ipsilateral breast cancer recurrence, 5) male sex, and 6) neoadjuvant chemotherapy.
Pathological primary tumor size was classified according the seventh edition of the AJCC Staging Manual.8 Cutoffs for estrogen receptor (ER) and progesterone receptor (PR) positivity were 10% positive cells, irrespective of intensity. Human epidermal growth factor receptor 2 (HER2) positivity was defined as either HER2 3+ (strongly positive in >10% of cancer cells on immunohistochemistry), or HER2 2+ (moderately positive in >10% of cancer cells) and gene amplification ≥2-fold on fluorescence in situ hybridization. Intrinsic subtypes were defined by ER, PR, and HER2 status: luminal subtype as ER+ and/or PR+, and HER2−; luminal/HER2 subtype as ER+ and/or PR+, and HER2+; HER2 subtype as ER−, PR−, and HER2+; and triple-negative subtype as ER−, PR−, and HER2−.
Sentinel Lymph Node Biopsy Procedure
The RI tracer used was 1.5 mCi/mL of 99mTc-phytate. One day before surgery, the tracer was injected into the intradermal and subdermal space in the area of the tumor and into the retrotumoral space. In all cases, lymphoscintigraphy was obtained 1 hour after the injection. In addition, 2 to 3 mL of vital dye, indigo carmine, was injected into the peritumoral space or areola at surgery. Before surgery, to resect the primary tumor, sentinel lymph nodes were identified using a handheld gamma-probe with the assistance of vessel and lymph node staining.
All sentinel lymph nodes were step-sectioned at 2 mm intervals by surgeons. Each of these sections was intraoperatively assessed by FS histology with hematoxylin and eosin stain. Immunohistochemistry was not used for the evaluation of sentinel lymph nodes.
We reviewed the FS specimens and classified them into 4 categories according to the seventh edition of the AJCC Staging Manual8: positive, macrometastasis (>2.0 mm); positive, micrometastasis (>0.2 mm to ≤2 mm); negative, ITCs (≤0.2 mm); and negative and no tumor cells. When cancer cells were observed in several sections or lymph nodes, the priority order of the sentinel lymph node status was macrometastasis, micrometastasis, and ITC.
The OSNA assay for lymph nodes was described in detail in a previous report.10 Briefly, after removing extranodal tissue and lipid, a whole lymph node was homogenized with 4 mL of a lysis buffer solution (Lynorhag; Sysmex) and centrifuged at 10,000 × g at room temperature. A 2 μL sample of the supernatant was analyzed with the RD-100i system (Sysmex), an automated gene amplification detection system using a reverse transcription loop-mediated isothermal amplification method,14 and with the LynoampBC (Sysmex). The degree of amplification was detected via a byproduct of the reaction, pyrophosphate.15 Upon precipitation of magnesium pyrophosphate, the resulting change in turbidity was in turn correlated to CK19 mRNA copy number/μL of the original lysate via a standard curve that was established beforehand with 3 calibrators containing different CK19 mRNA concentrations. A standard positive control sample containing 5000 copies/μL of CK19 mRNA and a negative control sample not containing any CK19 mRNA were used for validation in every assay. Lymph nodes that exceeded the specified maximum weight of 600 mg were cut into 2 or more pieces and processed as separate nodes. Four lymph nodes can be analyzed in 1 run.
The number of CK19 mRNA copies per microliter was calculated, and based on this number, the result (positive/negative) was assessed in accordance with the cutoff level determined by Tsujimoto et al.10 The OSNA assay can classify a positive lymph node into 3 categories: (++), (+), and (+I) (positive with reaction inhibited); criteria are shown in Table 1. In the present study, OSNA (++) and (+I) were considered to be equivalent to AJCC macrometastasis, and OSNA (+) to AJCC micrometastasis. Moreover, when multiple sections or lymph nodes were positive, the priority order of the sentinel lymph node status was (++), (+I), and (+).
Table 1. Definition of Positive Lymph Nodes by the One-Step Nucleic Acid Amplification Assay
|≥5000||Positive (++)||Positive (++)|
|250-5000||Positive (+)||Positive (+)|
Turnaround time was defined as the period from arrival of sentinel lymph nodes in our pathology department to getting the full results. The following situations were excluded from calculation of turnaround time: 1) time not recorded, 2) simultaneous submission with another patient's lymph nodes, and 3) submission of lymph nodes twice. When nonsentinel lymph nodes were also submitted for the intraoperative assay, the turnaround time included the assay of these nodes.
First, to compare the characteristics of patients and tumors that underwent sentinel lymph node biopsy between FS histology and OSNA assay, we performed the chi-square test for each characteristic. Second, to compare detection rates of sentinel lymph node metastases, including macro- and micrometastases, between FS histology and OSNA assay, we performed a 2-population z test for sentinel lymph node metastases (positive vs negative), macrometastases (macrometastasis by FS vs OSNA [++] or [+I]), and micrometastases (micrometastasis by FS vs OSNA [+]). Third, to reveal patient and tumor characteristics for which OSNA can detect metastases but FS cannot, we performed the Breslow-Day-Tarone test16, 17 for each characteristic. Computation of these analyses was performed by the statistical software R, version 220.127.116.11 A P value of <.05 was considered significant. Confidence intervals (CIs) were set at the 95% level.
Subjects comprised 618 patients in the FS cohort and 531 patients in the OSNA cohort. The demographic characteristics of the 2 cohorts are presented in Table 2. These 2 cohorts were well balanced for almost all tumor and patient characteristics except sentinel lymph node identification method (RI alone vs RI plus dye). In this regard, the proportion of patients in the OSNA cohort who had a combination of RI plus dye was higher than that in the FS cohort (OSNA 424 of 531, 79.8% vs FS 461 of 618, 74.6%; P = .035). The addition of dye to RI did not increase the yield of sentinel lymph nodes. Median number of identified sentinel lymph nodes was 2 irrespective of the addition of dye in both cohorts.
Table 2. Patient Characteristics of the Frozen Section Histology and OSNA Assay Cohorts
|No. of patients||618||100.0||531||100.0|| |
|Age, y|| || || || ||.53|
| Median (range)||54 (25-90)||55 (26-89)|| |
| ≤40||71||11.5||52||9.8|| |
| 41-69||471||76.2||419||78.9|| |
| ≥70||76||12.3||60||11.3|| |
|Menstrual status|| || || || ||.97|
| Premenopausal||267||43.2||232||43.7|| |
| Postmenopausal||351||56.8||299||56.3|| |
|Tumor location|| || || || ||.31|
| Upper inner||153||24.8||149||28.1|| |
| Lower inner||66||10.7||44||8.3|| |
| Upper outer||275||44.5||247||46.5|| |
| Lower outer||113||18.3||85||16.0|| |
| Central||11||1.8||6||1.1|| |
|Breast surgery|| || || || ||.42|
| Conservative||420||68.0||349||65.7|| |
| Mastectomy||198||32.0||182||34.3|| |
|SN identification|| || || || ||.035a|
| RI alone||157||25.4||107||20.2|| |
| RI plus dye||461||74.6||424||79.8|| |
|No. of SNs, median (range)|| || || || || |
| RI alone||2 (1-6)||2 (1-8)|| |
| RI plus dye||2 (1-6)||2 (1-10)|| |
|Pathological T classification|| || || || ||.27|
| pTmi||34||5.5||34||6.4|| |
| pT1a||87||14.1||83||15.6|| |
| pT1b||136||22.0||134||25.2|| |
| pT1c||262||42.4||191||36.0|| |
| pT2||99||16.0||89||16.8|| |
|Histological type|| || || || ||.60|
| Ductal||545||88.2||478||90.0|| |
| Lobular||17||2.8||13||2.4|| |
| Special types||56||9.1||40||7.5|| |
|Nuclear grade|| || || || ||.10|
| 1||261||42.2||197||37.1|| |
| 2||200||32.4||189||35.6|| |
| 3||84||13.6||92||17.3|| |
| Lobular/special type||73||11.8||53||10.0|| |
|Lymphovascular invasion|| || || || ||.91|
| −||458||74.1||395||74.4|| |
| +||160||25.9||136||25.6|| |
|Fat invasion|| || || || ||.059|
| −||151||24.4||156||29.4|| |
| +||467||75.6||375||70.6|| |
|Estrogen receptor status|| || || || ||.52|
| −||105||17.0||98||18.5|| |
| +||513||83.0||433||81.5|| |
|Progesterone receptor status|| || || || ||.48|
| −||249||40.3||203||38.2|| |
| +||369||59.7||328||61.8|| |
|HER2 status|| || || || ||.55|
| −||544||88.0||457||86.1|| |
| +||64||10.4||62||11.7|| |
| Unknown||10||1.6||12||2.3|| |
|Intrinsic subtype|| || || || ||.80|
| Luminal||481||77.8||409||77.0|| |
| Luminal/HER2||29||4.7||26||4.9|| |
| HER2||35||5.7||36||6.8|| |
| Triple negative||63||10.2||48||9.0|| |
| Unknown||10||1.6||12||2.3|| |
Detection of Sentinel Lymph Node Metastasis
Sentinel lymph nodes were found to be positive for metastasis significantly more frequently in the OSNA cohort than in the FS cohort (FS 109 of 618, 17.6%, 95% CI, 14.8%-20.9% vs OSNA 121 of 531, 22.8%, 95% CI, 19.3%-26.6%; P = .036) (Table 3). We found no significant difference in the frequency of histological macrometastasis with OSNA (++) or (+I) in sentinel lymph nodes (FS 81 of 618, 13.1%, 95% CI, 10.6%-16.1% vs OSNA 75 of 531, 14.1%, 95% CI, 11.3%-17.4%; P = .68). Conversely, we found significant difference in the frequency of histological micrometastasis and OSNA (+) in sentinel lymph nodes (FS 28 of 618, 4.5%, 95% CI, 3.1%-6.6% vs OSNA 46 of 531, 8.7%, 95% CI, 6.5%-11.5%; P = .0064).
Table 3. Detection of Sentinel Node Metastasis in the Frozen Section Histology and OSNA Assay Cohorts
| Macrometastasis||81||13.1%||10.6-16.1|| (++) or (+I)||75||14.1%||11.3-17.4||.68|
| Micrometastasis||28||4.5%||3.1-6.6|| (+)||46||8.7%||6.5-11.5||.0064a|
| Isolated tumor cells||9||1.5%|| || || || || || |
| No tumor cell||500||80.9%|| || || || || || |
Characteristics of Patients for Whom the OSNA Assay Could Detect Sentinel Lymph Node Metastases But FS Histology Could Not
Compared with FS histology, the OSNA assay could detect more sentinel lymph node metastases in patients who were aged >40 years (FS 90 of 547, 16.5% vs OSNA 112 of 479, 23.4%; P = .036) or postmenopausal (FS 43 of 351, 12.3% vs OSNA 77 of 302, 25.5%; P < .0001), or from tumors without fat invasion (FS 6 of 151, 4.0% vs OSNA 23 of 156, 14.7%; P = .012) or without lymphovascular invasion (FS 45 of 458, 9.8% vs OSNA 67 of 395, 17.0%; P = .042) (Table 4). Furthermore, the OSNA assay tended to detect more sentinel lymph node metastases from microinvasive tumors (pTmi and pT1a) than did FS histology (FS 3 of 121, 2.5% vs OSNA 12 of 117, 10.3%; P = .057). Method of sentinel lymph node identification was not a significant factor regarding whether the OSNA assay was positive.
Table 4. Characteristics of Patients for Whom OSNA Assay But Not Frozen Section Histology Could Detect SN Metastases
|No. of patients||618||109||17.6%||531||121||22.8%|| |
|Age, y|| || || || ||.036ab|
| ≤40||71||19||26.8%||52||9||17.3%|| |
| 41-69||471||82||17.4%||419||100||23.9%|| |
| ≥70||76||8||10.5%||60||12||20.0%|| |
|Menstrual status|| || || || || || ||<.0001a|
| Premenopausal||267||66||24.7%||232||45||19.4%|| |
| Postmenopausal||351||43||12.3%||299||76||25.4%|| |
|Tumor location|| || || || || || ||.53c|
| Upper inner||153||23||15.0%||149||30||20.1%|| |
| Lower inner||66||12||18.2%||44||13||29.5%|| |
| Upper outer||275||49||17.8%||247||57||23.1%|| |
| Lower outer||113||22||19.5%||85||18||21.2%|| |
| Central||11||3||27.3%||6||3||50.0%|| |
|Breast surgery|| || || || || || ||.46|
| Conservative||420||67||16.0%||349||67||19.2%|| |
| Mastectomy||198||42||21.2%||182||54||29.7%|| |
|SN identification|| || || || || || ||.82|
| RI alone||157||24||15.3%||107||20||18.7%|| |
| RI plus dye||461||85||18.4%||424||101||23.8%|| |
|Pathological T classification|| || || || || || ||.057d|
| pTmi||34||0||0%||34||4||11.8%|| |
| pT1a||87||3||3.4%||83||8||9.6%|| |
| pT1b||136||14||10.3%||134||19||14.2%|| |
| pT1c||262||60||22.9%||191||57||29.8%|| |
| pT2||99||32||32.3%||89||33||37.1%|| |
|Histological type|| || || || || || ||.43e|
| Ductal||545||97||17.8%||478||111||23.2%|| |
| Lobular||17||4||23.5%||13||5||38.5%|| |
| Special types||56||8||14.3%||40||5||12.5%|| |
|Nuclear grade|| || || || || || ||.12f|
| 1||261||33||12.6%||197||42||21.3%|| |
| 2||200||49||24.5%||189||47||24.9%|| |
| 3||84||15||17.9%||92||22||23.9%|| |
| Lobular/special type||73||12||16.4%||53||10||18.9%|| |
|Lymphovascular invasion|| || || || || || ||.042a|
| −||458||45||9.8%||395||67||17.0%|| |
| +||160||64||40.0%||136||54||39.7%|| |
|Fat invasion|| || || || || || ||.012a|
| −||151||6||4.0%||156||23||14.7%|| |
| +||467||103||22.1%||375||98||26.1%|| |
|Estrogen receptor status|| || || || || || ||.62|
| −||105||15||14.3%||98||21||21.4%|| |
| +||513||94||18.3%||433||100||23.1%|| |
|Progesterone receptor status|| || || || || || ||.80|
| −||249||42||16.9%||203||46||22.7%|| |
| +||369||67||18.2%||328||75||22.9%|| |
|HER2 status|| || || || || || ||.89g|
| −||544||94||17.3%||457||101||22.1%|| |
| +||64||15||23.4%||62||19||30.6%|| |
| Unknown||10||0||0%||12||1||8.3%|| |
|Intrinsic subtype|| || || || || || ||.27h|
| Luminal||481||88||18.3%||409||92||22.5%|| |
| Luminal/HER2||29||8||27.6%||26||8||30.8%|| |
| HER2||35||7||20.0%||36||11||30.6%|| |
| Triple negative||63||6||9.5%||48||9||18.8%|| |
| Unknown||10||0||0%||12||1||8.3%|| |
Turnaround Time of OSNA Assay
Turnaround time could be assessed in 461 patients after the exclusion of 70 patients. Median number of specimens was 3 (range, 1-12). Median and mean turnaround times were 41 and 43 minutes (range, 25-116 minutes), respectively.
As far as we know, this is the first clinical report to input a whole sentinel lymph node in a molecular assay. Although this is a retrospective cohort study, both cohorts were relatively large and well balanced for patient characteristics. The present study shows that the OSNA assay using a whole lymph node can detect more sentinel lymph node metastases, particularly micrometastases, than routine FS histology using a 2 mm-sectioned lymph node. The OSNA assay increased the sentinel lymph node-positive rate by 30% from the FS histology. These results are reasonable, because 2 mm-section histology can miss metastases <2 mm between the cut surfaces, and the false-negative rate of intraoperative FS histology, mainly resulting from failure to detect micrometastases, has been found to range from approximately 30% to 40% compared with the final pathological results.9
The OSNA assay could detect more sentinel lymph node metastases than could FS histology from primary tumors without lymphovascular or fat invasion, or from microinvasive tumors. Several studies have shown that primary tumors with nodal micrometastases are smaller than those with macrometastases.19-22 Therefore, the majority of lymph node metastases from these less advanced tumors can be expected to be micrometastases, leading to false-negative cases with FS histology. Furthermore, the OSNA assay could detect more metastases from elderly or postmenopausal patients than FS histology. A U-shaped or negative linear relation has been reported between age and nodal metastasis.23, 24 The results of the present OSNA cohort could not replicate the abovementioned relationship, although the FS cohort showed the negative linear relationship. However, the clinical significance of this finding is unclear.
The turnaround time of the OSNA assay was approximately 40 minutes, which is slightly longer than that for routine FS histology (<30 minutes). However, this time is feasible in daily clinical practice, because surgeons can perform the excision of the primary tumor by partial or total mastectomy while the assay is running. Although the relative time efficiency of the assay is important, the primary merit of the OSNA assay is the finding that an exhaustive evaluation of the whole lymph node is possible with high accuracy.
Although the OSNA assay detects more sentinel lymph node micrometastases than FS histology, the clinical impact of sentinel lymph node micrometastasis continues to be an area of great debate. The risk of further nonsentinel lymph node involvement was reported to be approximately 15% to 20% in patients with sentinel lymph node micrometastasis detected by histological examination25-29; however, it remains controversial whether completion axillary lymph node dissection is needed for patients with sentinel lymph node micrometastasis30-34 or micrometastasis has a prognostic impact.2, 19, 21, 35, 36 One of the reasons for these controversies is that the conventional histopathological examination is not standardized, and its ability to measure accurate total metastatic volume in a lymph node, particularly for low-volume metastases, is limited. In contrast, because the OSNA assay can evaluate a whole lymph node in a standardized manner, follow-up of the OSNA cohort can reveal the clinical impact of sentinel lymph node micrometastasis. Furthermore, in addition to sentinel lymph nodes, assessing all the additional nonsentinel lymph nodes by OSNA assay can show the total metastatic volume of all axillary nodes. Therefore, follow-up of the patients with true micrometastasis can clarify the prognostic impact with greater accuracy. As a result, it may be possible to establish a new nodal staging system using OSNA results.
In terms of the AJCC staging system, pN classification is based on the results of histopathological examination. The correspondence between the results of molecular assay and histopathological metastatic size is limited, particularly when a whole lymph node is used for the molecular assay. We defined OSNA (++) and (+I) as equivalent to AJCC macrometastasis, and OSNA (+) as equivalent to AJCC micrometastasis in the present study. Theoretically, OSNA (++) and (+) can be defined as histological macro- and micrometastasis, respectively. According to permanent 3-level CK19 immunohistochemistry-based histopathology, the cutoff value of 5000 copies/μL in the OSNA assay is 85% concordant with macrometastasis.10 On the basis of a serial sectioning experiment, moreover, the cutoff value of 250 copies/μL in the OSNA assay can detect metastatic foci of 0.3 to 0.4 mm3.10 However, there is a lack of evidence regarding OSNA (+I) defined as macrometastasis. The reverse transcription loop-mediated isothermal amplification method is occasionally inhibited by unknown material, but this inhibition is mitigated in the diluted samples. We considered OSNA (+I) as indicative of high-volume metastasis corresponding to AJCC macrometastasis, because CK19 mRNA was ≥250 copies/μL in the 10-fold more diluted samples. However, the small number of OSNA (+I) cases probably included some with low-volume metastasis corresponding to AJCC micrometastasis. In the future, we hope the development of a reagent that does not inhibit the reaction will allow classification according to accurate metastatic volume.
The incidence of false positives appears to be small but considerable for the OSNA assay. The OSNA assay cannot discriminate between benign and malignant epithelial cells showing CK19-positivity. Therefore, it is possible that lymph nodes with contamination of benign epithelial cells, benign transport of breast epithelial cells, or benign intranodal inclusions cause false-positive diagnoses. In terms of the contamination, the rigorous sentinel lymph node biopsy with RI tracer and the removal of extranodal tissue before homogenization minimizes false-positive diagnoses in the present study. Moreover, benign transport of breast epithelial cells can be excluded for the false positives, because almost all benign epithelial cells in sentinel lymph nodes are small clusters corresponding to ITCs.37 Because the OSNA assay is designed not to recognize tumor cells corresponding to ITCs, this is unlikely to be a source of error in the test. However, benign intranodal epithelial inclusions, such as heterotopic mammary glands, benign glandular inclusions, or benign Mullerian inclusions, are inevitable for the false-positive diagnoses,38-41 although their presence is a rare event in axillary lymph nodes. Clinical trials comparing OSNA assay and histopathological examination have demonstrated a high specificity (96% to 97%) after investigating discordant cases.11-13
The incidence of false negatives also appears to be very low for the OSNA assay. The OSNA assay can miss metastases that do not express CK19 mRNA. Approximately 98% of primary breast cancer tissue express the CK19 protein.10, 42 Considering the sentinel lymph node-positive rate, approximately 0.5% of patients undergoing sentinel lymph node biopsy are likely to have a false-negative result. Moreover, it is possible that breast cancer tissues express CK19 mRNA despite the low protein expression found by CK19 immunohistochemistry.13
In conclusion, the intraoperative OSNA assay detects more sentinel lymph node metastases, particularly micrometastases, than FS histology. The OSNA assay can also detect more metastases in elderly patients or from less progressive primary tumors than FS histology. Moreover, the turnaround time of the assay is feasible in daily clinical practice. Follow-up of patients who have undergone the OSNA assay is required to clarify the prognostic implications; this may lead to the establishment of a new breast cancer staging using OSNA results, contributing to personalized medicine.
We thank M. Makita, S. Nishimura, Y. Miyagi, K. Iijima, T. Morizono, T. Sakai, K. Masumura, R. Gokan, Y. Ishikawa, K. Takeuchi, T. Miyagi, T. Kiniwa, K. Kishi, and K. Iwakabe.
This work was supported by a Grant-in-Aid for Young Scientist (B) (No. 21791264) from the Japanese Ministry of Education, Culture, Sports, Science, and Technology.
CONFLICT OF INTEREST DISCLOSURES
The authors made no disclosures.