Clinical and Translational Oncology (2021) 23:1377–1385 https://doi.org/10.1007/s12094-020-02530-4 RESEARCH ARTICLE Predictive and prognostic value of total tumor load in sentinel lymph nodes in breast cancer patients after neoadjuvant treatment using one‑step nucleic acid amplification: the NEOVATTL study B. Vieites1 · M. Á. López‑García1,11 · M. D. Martín‑Salvago2 · C.L. Ramirez‑Tortosa2 · R. Rezola3 · M. Sancho4 · L. López‑Vilaró5 · F. Villardell6 · O. Burgués7 · B. Fernández‑Rodriguez8 · L. Alfaro9 · V. Peg10,11 Received: 8 October 2020 / Accepted: 18 November 2020 / Published online: 31 January 2021 © The Author(s) 2021 Abstract Objective To evaluate the predictive and prognostic value of total tumor load (TTL) in sentinel lymph nodes (SLNs) in patients with infiltrating breast cancer after neoadjuvant systemic therapy (NST). Methods This retrospective multicenter study used data from a Spanish Sentinel Lymph Node database. Patients underwent intraoperative SLN biopsy after NST. TTL was determined from whole nodes using a one-step nucleic acid amplification (OSNA) assay and defined as the total sum of CK19 mRNA copies in all positive SLNs. Cox-regression models identified independent predictive variables, which were incorporated into a nomogram to predict axillary non-SLN metastasis, and identified prognostic variables for incorporation into a disease-free survival (DFS) prognostic score. Results A total of 314 patients were included; most had no lymph node involvement prior to NST (cN0; 75.0% of patients). Most received chemotherapy with or without biologic therapy (91.7%), and 81 patients had a pathologic complete response. TTL was predictive of non-SLN involvement (area under the concentration curve = 0.87), and at a cut-off of 15,000 copies/ µL had a negative predictive value of 90.5%. Nomogram parameters included log (TTL + 1), maximum tumor diameter and study-defined NST response. TTL was prognostic of disease recurrence and DFS at a cut-off of 25,000 copies/µL. After a 5-year follow-up, DFS was higher in patients with ≤ 25,000 copies/µL than those with > 25,000 (89.9% vs. 70.0%; p = 0.0017). Conclusions TTL > 15,000 mRNA copies/µL was predictive of non-SLN involvement and TTL > 25,000 mRNA copies/µL was associated with a higher risk of disease recurrence in breast cancer patients who had received NST. Keywords Breast cancer · Disease-free survival · Neoadjuvant systemic therapy · OSNA · Sentinel lymph node · Total tumor load 6 * B. Vieites Hospital Universitario Arnau de Vilanova, Lérida, Spain mb.vieites.sspa@juntadeandalucia.es 7 Department of Pathology, Hospital Clínico Universitario de 1 Valencia, Valencia, Spain Department of Pathology, Hospital Universitario Virgen del 8 Rocío, Sevilla, Spain Department of Pathology, Complejo Hospitalario, 2 Universitario de Santiago, Santiago de Compostela, Spain Department of Pathology, Hospital Universitario 9 Materno-Infantil, Jaén, Spain Department of Gynaecology and Obstetrics, Hospital 3 Universitario Virgen del Rocío, Sevilla, Spain Department of Pathology, Onkologikoa Kutxa Fundazioa, 10 Donostia, Spain Department of Pathology, Hospital Universitari Vall 4 d’Hebron, Barcelona, Spain Department of Pathology, Hospital Universitario de 11 Salamanca, Salamanca, Spain CIBERONC (Centro de Investigación Biomédica en Red de 5 Cáncer), Instituto de Salud Carlos III, Madrid, Spain Department of Pathology, Hospital de La Santa Creu I Sant Pau, Barcelona, Spain 13 Vol.:(0123456789) 1378 Clinical and Translational Oncology (2021) 23:1377–1385 Introduction Methods In 2018, breast cancer was the most common cancer in Study population women worldwide, with 2.1 million incident cases, rep- resenting 25.4% of all cancers in women [1]. With the A historical cohort of patients diagnosed with infiltrating drive for breast-conserving surgery, as well as early treat- breast cancer between 2009 and 2012 was selected for this ment for presumed micrometastatic disease, there is an study. Patients where included if they had received NST increasing number of early-stage breast cancer patients and had undergone SLNB and evaluation of all complete treated with neoadjuvant systemic therapy (NST) prior to excised SLNs by OSNA. Patients were excluded from this surgery [2, 3]. Assessment of regional lymph node status analysis if they were diagnosed with carcinoma in situ provides important prognostic information, and in the case without an infiltrating component or any other concomi- of patients who have undergone NST, sentinel lymph node tant neoplasia. biopsy (SLNB) has gained attention because of its poten- Due to the retrospective nature of this study, selection of tial for less invasive management of the axilla; it should patients for SLNB was per usual institutional practice, i.e. ideally take place after, not prior to, NST [4]. One-step patients with minimal or null node involvement evaluated nucleic acid amplification (OSNA) assay is an accurate by imaging after NST and deemed to be candidates who and reliable option for intraoperative molecular analysis would benefit from avoiding axillary dissection. Likewise, of SLN status [5]. NST was administered according to the treating oncologists’ The OSNA assay (Sysmex, Kobe, Japan) analyzes the decision and per institutional protocols. whole SLN through a semi-quantitative result based on the detection and real-time reverse transcription-loop-medi- ated isothermal amplification (RT-LAMP) of cytokeratin Data and data source 19 (CK19) mRNA, and provides accurate intraoperative detection of lymph node metastases [6–11]. High con- Patient data were from eight Spanish hospitals on the cordance between OSNA and conventional techniques in LYNOLOG database, an online database created by the the detection of breast cancer cells has been observed [6, Spanish Society for the Study of Sentinel Node (Sociedad 12]. OSNA is a commonly used axillary staging method Española para el Estudio del Ganglio Centinela, SEEGC). in Europe and Asia, but there has been debate on its use in The Spanish Society of Senology and Breast Disease (Socie- patients who have undergone NST [5]. There is, however, dad Española de Senología y Patología Mamaria, SESPM) evidence that CK19 expression is preserved in breast can- gave permission to use this database and the Spanish Society cer cells after NST [12, 13] and some studies have shown of Anatomy and Pathology (Sociedad Española de Anatomía identification of SLN metastases in patients following NST Patológica, SEAP) acted as the guarantor. The study was is similar to that in patients who have not undergone NST conducted in accordance with the Declaration of Helsinki [14, 15]. and guidelines for Good Clinical Practice, after obtaining In addition, the role of total tumor load (TTL; amount PEIBA Ethics Committee (Portal de Ética de la Investi- of CK19 mRNA copies in all positive SLN) for assessing gación Biomédica de Andalucía) approval. non-SLN involvement and patient prognosis continues to Clinical, pathologic and treatment data were collected, be an area of active investigation in early breast cancer including patient characteristics (sex), NST (type, start [16, 17], with a continued focus on sparing patients from and end date), radiotherapy received (yes/no, location axillary lymph node dissection (ALND). However, data on [breast ± axilla]), tumor characteristics (histologic subtype the predictive and prognostic value of TTL derived from and grade, presence of in situ carcinoma, lymphovascular intra-operative OSNA assay in breast cancer patients after invasion, hormone receptor status, human epidermal growth NST are lacking. The aim of this retrospective study was to factor receptor 2 [HER2] status, Ki67 proliferation index evaluate the predictive and prognostic value of TTL deter- score, TNM stage before and after NST [cTNM and ycTNM, mined from OSNA assessment of SLNs in the neoadjuvant respectively]), and histologic tumor response according to setting in breast cancer patients. the Miller–Payne grading system [18] (see Supplementary Methods in Online Resource 1 for a complete list). Pre‑operative sentinel node assessment Axillary lymph nodes (ALNs) were clinically evaluated per institutional protocol, i.e. usually by clinical and ultrasound 13 Clinical and Translational Oncology (2021) 23:1377–1385 1379 examination. Any clinically suspicious ALN underwent Area under the receiver operating characteristic (ROC) histologic confirmation by fine-needle aspiration or core curve (AUC) was determined to evaluate the predictive biopsy. SLNB was performed if ≤ 1 ALN was positive for value of TTL for axillary non-SLN involvement. metastasis; in general, ALND was performed if some ALNs Cox-regression models were constructed to assess were positive (with no SLN analysis undertaken). the main study objectives. Univariate and multivariate analyses were conducted to identify factors predictive Detection and OSNA assay of SLNs of non-SLN involvement, from which a nomogram was constructed to predict non-SLN involvement. For assess- Each institution’s pathology department examined each ment of prognostic factors, an initial univariate analysis whole lymph node using the OSNA assay, composed of the included time to event (recurrence) as the dependent vari- automated Gene Amplification Detector RD-100i and the able, and all potential prognostic factors, including TTL, LYNOAMP BC gene amplification reagent (Sysmex Corpo- as independent variables. Thereafter, independent vari- ration, Kobe, Japan). The assay technique has been described ables reaching a p-value < 0.25 were included in the mul- in detail previously [10]. Briefly, the SLN was homogenized tivariate analysis. DFS was calculated by Kaplan–Meier in 4 mL of LYNORHAG lysis buffer (Sysmex Corporation, survival analysis. A score to estimate 5-year DFS was also Kobe, Japan). A 2 µL aliquot was used for automated real- calculated (Supplementary Methods—Online Resource 1). time amplification of CK19 mRNA with the ready-to-use The planned sample size was 296 patients, determined LYNOAMP reagent on the RD-100i analyzer. Lymph node using estimates from previous studies (disease recurrence metastatic status was determined according to the manufac- rate of 16% [21] and SD of 1.3 [22]), and calculated using turer’s criteria: metastatic (OSNA positive) ≥ 2.5 × 102 CK19 PASS software version 13 (NCSS Statistical Software). mRNA copies/µL (hereafter referred to as copies/µL), and This would provide a statistical power of 90% with an non-metastatic (OSNA negative) < 2.5 × 102 copies/µL [10]. alpha-error of 5%. Statistical analyses were performed using SAS software version 9.4 (SAS Institute Inc., Cary, Study definitions NC, USA). TTL was defined as the total sum of CK19 mRNA copies in all positive SLN, and TTL cut-offs defined per Peg et al. [19]. In this analysis, pathologic complete response to NST was defined as a lack of invasive disease according to the TNM approach in the AJCC Cancer Staging Manual (8th Table 1  Patient demographic and clinical characteristics ed) [20], i.e. as no invasive breast tumor (by histopathology) Characteristic All included and no metastases in the nodes (by OSNA assay of SLN or patients histologic analysis of non-SLNs). Treatment response was (n = 314) evaluated histologically using the Miller–Payne grading sys- At diagnosis tem. Disease-free survival (DFS) was defined as the period Gender of time with no evidence of disease. Male, n (%) 1 (0.3) Female, n (%) 313 (99.7) Study objective Tumor type, n (%) Ductal 262 (84.0) The primary objectives were to determine the predictive Other 50 (16.0) value of TTL for the diagnosis of non-SLN metastases in Neoadjuvant systemic therapy patients after NST and to determine the prognostic value of Duration, mean ± SD (months) 4.8 ± 2.2 TTL in determining 5-year DFS. Number of cycles, mean ± SD 8.7 ± 3.3 Type, n (%) Statistics Chemotherapy ± biologic therapy 288 (91.7) Chemotherapy + hormonal 6 (1.9) Descriptive statistics were used for quantitative variables Hormonal therapy ± biologic therapy 20 (6.4) (mean, standard deviation [SD], median, interquartile range Adjuvant radiotherapy, n (%) [IQR, P25–P75], minimum and maximum) and categorical Breast ± axillary nodes 244 (77.7) variables (absolute or relative frequency with corresponding Breast + axillary nodes 50 (15.9) 95% confidence interval [CI]). Breast + axillary nodes + ALND 31 (9.9) The sensitivity, specificity, positive and negative predic- tive values, accuracy and error rate of TTL was calculated. ALND axillary lymph node dissection, SD standard deviation 13 1380 Clinical and Translational Oncology (2021) 23:1377–1385 Results Table 2  Pathologic tumor characteristics before and after neoadju- vant systemic therapy (NST) Patients and treatment Maximum tumor size Before NST After NST n = 267 n = 309 The study cohort consisted of 314 patients, of whom only ≤ 20 mm 25 (9.4) 125 (40.5) one was male (Table 1). Prior to NST, most patients did > 20 mm 242 (90.6) 103 (33.3) not have lymph node involvement (cN0; 75.0% of patients Histologic subtype n = 312 n = 312 assessed), 24.0% were classified as cN1 and 1.0% as cN2 Ductal (including NOS) 262 (84.0) 182 (58.3) (Table 2). Of the 231 cN0 patients, 129 underwent ALN Others 50 (16.0) 49 (15.7) histologic examination prior to NST, of whom 81 (25.8%) Histologic grade n = 296 n = 307 had no metastatic nodal involvement and 48 (15.3%) had G1 36 (12.2) 62 (20.2) metastatic ALN (but most with signs of remission after G2 and G3 260 (87.8) 164 (53.4) NST); 3 of these 48 patients underwent a SLNB both before In situ carcinoma n = 314 n = 261 and after NST. No 250 (79.6) 162 (62.1) Neoadjuvant ST for most patients was chemotherapy with Yes 64 (20.4) 99 (37.9) or without biologic or anti-human epidermal growth fac- Lymphovascular infiltration n = 314 n = 286 tor receptor-2 (HER2) agents (91.7%), whereas only 6.4% No 199 (63.4) 192 (67.1) of patients received hormone therapy (with or without bio- Yes 115 (36.6) 94 (32.9) logic/HER2 treatment) (Table 1). Adjuvant breast irradiation ER n = 311 n = 111 was administered to the majority of patients (n = 244), of Negative 84 (27.0) 21 (18.9) whom 50 also received ALN irradiation. Thirty-one patients Positive (≥ 1%) 227 (73.0) 90 (81.1) underwent both ALND and axillary irradiation; of these, 15 PR n = 305 n = 111 had positive SLNs, but no further axillary treatment was Negative 116 (38.0) 55 (49.5) performed. The maximum TTL for these cases was 6,700 Positive (≥ 1%) 189 (62.0) 56 (50.5) copies/µL. HER2 n = 314 n = 293 Negative 233 (74.2) 278 (94.9) Positive (≥ 1%) 81 (25.8) 15 (5.1) Histopathologic outcomes Ki67 n = 310 n = 107 > 20% 184 (59.4) 35 (32.7) After NST, 69 of 309 patients (22.3%) had a pCR. More ≤ 20% 126 (40.6) 72 (67.3) patients had a maximum tumor size of ≤ 20 mm and better TNM stage cTNM ycTNM histologic differentiation and staging after NST than before, 0.T n = 314 n = 309 and more patients were ycN1 after NST than those who were In situ NA 5 (1.6) designated cN1 prior to NST (30.3% vs. 24.0%; Table 2). 0 NA 77 (24.9) Pathologic tumor features pre- versus post-NST by TTL are 1 26 (8.3) 115 (37.2) shown in Table S1 in Online Resource 2. 2 228 (72.6) 94 (30.4) During surgery (after NST), OSNA was performed in all 3 52 (16.6) 17 (5.5) patients for molecular analysis of SLNs, and ALND added 4 8 (2.5) 1 (0.3) in 127 (40.4%) for whom OSNA revealed positive SLNs. N n = 308 n = 314 Pathologic and molecular analysis results are shown in 0 231 (75.0) 194 (61.8) Table 3. A total of 196 patients (62.4%) were considered to 1 74 (24.0) 95 (30.3) have negative results for residual cancer burden in axillary 2 3 (1.0) 24 (7.6) nodes (TTL < 250 mRNA copies/µL). During surgery, one 3 – 1 (0.3) or two SLN were removed in 227 (72.3%) patients whereas M n = 312 n = 314 87 patients (27.7%) had more than two removed (Table 3). 0 312 (100.0%) 314 (100.0) All data are n (%) unless otherwise specified Patient outcomes by TTL G grade, TTL total tumor load, pCR pathologic complete response, ER estrogen receptor, HER2 human epidermal growth factor recep- Median follow-up after surgery was 5.2 (IQR 1.5) years. A tor-2, PR progesterone receptor, cTNM clinical TNM tumor stage, ycTNM TNM stage after neoadjuvant therapy total of 17 (5.4%) patients died during follow up, of whom 7 (41.2%) had a TTL of > 250 copies/µL; 5 (29.4%) had a TTL of 250–25,000 copies/µL, and 5 (29.4%) had a TTL 13 Clinical and Translational Oncology (2021) 23:1377–1385 1381 Table 3  Clinicopathologic characteristics of patients derived from µL, 5,000 copies/µL and 15,000 copies/µL, respective val- histologic and molecular analyses ues for sensitivity were 95.7%, 63.0% and 43.5%, and for Variables specificity were 72.4%, 86.2% and 92.9%, positive predictive values were 37.3%, 43.9% and 51.3%, and negative predic- Total tumor load, copies/µL n = 314 tive values were 99.0%, 93.1% and 90.5%. Mean ± SD 42,314 ± 248,208 Univariate and multivariate logistic regression analyses 95% CI 14,754–69,874 were performed to investigate the factors predictive of non- Median (min–max) 0 (0–3,400,000) SLN metastasis (Table 4). Independent predictive factors Time since diagnosis, months n = 314 significantly associated with non-SLN metastasis were: TTL Mean ± SD 6.7 ± 1.6 ­(log10(TTL + 1)), treatment response (defined as patients 95% CI 6.5–6.8 with HER2-negative status and Ki67 ≤ 20% from needle Median (min–max) 6.6 (0.0–11.8) biopsy), and maximum tumor diameter (p = 0.0206, 0.0069 No. of SLN ­exciseda, n (%) n = 314 and 0.0080, respectively). The nomogram for prediction of 1 or 2 227 (72.3) non-SLN status is shown in Fig. 1. >2 87 (27.7) No. of metastatic ­SLNa, n (%) n = 314 Prognostic factors for recurrence and DFS 0 196 (62.4) 1–3 111 (35.4) Variables reaching significance of p<0.25 in a univariate >3 7 (2.2) analysis of factors prognostic of disease recurrence were ALND, n (%) n = 314 included in a multivariate analysis (data not shown). In the No 187 (59.6) multivariate analysis, independent predictors of disease Yes 127 (40.4) recurrence were TTL, Ki67 before NST, Miller–Payne grad- No. of non-SLN excised, n (%) n = 314 ing, and cN status. A TTL of >25,000 copies/μL was asso- <9 219 (69.7) ciated with a threefold risk of recurrence compared with a ≥9 95 (30.3) TTL of ≤ 25,000 copies/μL (hazard ratio [HR] 2.95; 95% CI No. of metastatic non-SLN, n (%) n = 314 1.39–6.28). Similarly, the risk of recurrence in patients with 0 268 (85.4) a Ki67 score of >20% prior to NST was more than fourfold 1–3 23 (7.3) compared with those with a Ki67 of ≤ 20% (HR 4.35; 95% >3 23 (7.3) CI 1.86–10.16). A reduction in treatment response by one Tumor pathologic response, n (%) n = 301 Miller–Payne grade was associated with a 52% higher risk Grade 1 22 (7.3) of recurrence (HR 1.52; 95% CI 1.16–2.00). There was a Grade 2 42 (14.0) twofold higher risk of disease recurrence associated with Grade 3 114 (37.9) an increase in cN status by one stage (HR 2.10; 95% CI Grade 4 38 (12.6) 1.17–3.75). Grade 5 85 (28.2) There was no significant difference in DFS between a At the time of the SLNB after NST patients with a TTL of < 250 copies/µL and those with a TTL ALND axillary lymph node dissection, CI confidence interval, no. of 250–25,000 copies/µL (Fig. 2a). However, DFS rates were number, SD standard deviation, SLN sentinel lymph node significantly higher in patients with a TTL of ≤ 25,000 cop- ies/µL compared with those patients with a TTL of > 25,000 copies/µL (89.9% vs. 70.0%, p = 0.0017; Fig. 2a). Adjust- of > 25,000 copies/µL. Distant metastases were detected in ment to remove the effect of Ki67 before NST, Miller–Payne 35 patients (11.1%), of whom 16 (45.7%), 9 (25.7%) and 10 grading and cN status resulted in an even lower DFS for (28.6%) were in each TTL category, respectively. Respective patients with a TTL of > 25,000 copies/µL (Fig. 2b). disease recurrence rates were 51.3%, 23.1%, and 25.6%, with A score to estimate 5-year DFS was derived, where DFS an overall disease recurrence rate of 12.4% (39 patients). decreased with a high TTL (≥ 25,000 copies/µL), high Ki67 (> 20%), or low Miller–Payne grade (Grade 1 or 2; Fig. 2c). Predictive value of TTL and clinicopathologic factors for non‑SLN involvement Discussion The ROC curve analysis demonstrated univariate concord- ance between TTL and non-SLN involvement (Online Our study confirmed the predictive and prognostic value Resource 3, Fig. S1), with an AUC of 0.87 (95% CI of TTL in breast cancer patients who had undergone sur- 0.823–0.918). When the TTL-cutoff was fixed at 250 copies/ gery and SLNB after NST. TTL had stand-alone predictive 13 1382 Clinical and Translational Oncology (2021) 23:1377–1385 Table 4  Summary of univariate and multivariate analysis for non-SLN metastasis Variable Univariate analysis Multivariate analysis OR (95% CI) p value OR (95% CI) p value Clinical Max. tumor diameter > 20 mm vs. ≤ 20 mm 3.23 (1.52–6.86) 0.0023 3.12 (1.34–7.22) 0.0080 Tumor grade G1 vs. pCR 13.81 (1.62–117.94) 0.0323 G2/ G3 vs. pCR 11.97 (1.51–94.69) 0.0494 LVI 2.47 (1.15–5.27) 0.0197 Yes vs. No Pathologic Log10(TTL + 1) 1.54 (1.20–1.96) 0.0006 1.36 (1.05–1.75) 0.0206 TTL > 25,000 vs. ≤ 25,000 copies/µL 3.37 (1.48–7.66) 0.0038 HER2 status Positive vs. Negative 0.26 (0.08–0.80) 0.0186 Ki67 > 20% vs. ≤ 20% 0.31 (0.14–0.66) 0.0025 Treatment response HER2-negative and Ki67 ≤ 20% from 3.22 (1.38–7.54) 0.0069 needle biopsy Yes vs. No 3.58 (1.68–7.66) 0.0010 Miller–Payne ­responsea 0.69 (0.48–0.98) 0.0382 a Comparison among grades CI confidence interval, HER2 human epidermal growth factor receptor 2, LVI lymphovascular infiltration, OR odds ratio, max. maximum, pCR pathologic complete response, SLN sentinel lymph node, TTL total tumor load Fig. 1  Nomogram to predict axillary non-SLN involve- ment in breast cancer patients after neoadjuvant systemic therapy. Probability refers to the probability of the presence of non-SNL metastasis. HER2, human epidermal growth factor receptor-2; Ki67, proliferation index; SLN, sentinel lymph node, TTL, total tumor load value for non-SLN metastases (AUC = 0.87). A nomogram independent prognostic factor for disease recurrence and was constructed, based on TTL and clinicopathologic fac- DFS over 5-year follow-up period. tors also identified as predictive of non-SLN involvement, Previous studies showing that TTL determined from and included a tumor size > 20 mm, and NST treatment molecular analysis is an independent predictor of metastatic response, and HER2-negative status in combination with non-SLN in breast cancer patients differed markedly from a Ki67 ≤ 20%. TTL at a cut-off of 25,000 copies/µL was an ours by having excluded patients who had received NST 13 Clinical and Translational Oncology (2021) 23:1377–1385 1383 Fig. 2  Kaplan–Meier estimates of 5-year disease-free survival, a copies/µL, a Ki67 of ≤ 20% or Miller–Payne grade of 5, Score 2 stratified by TTL (TTL < 250 copies/µL [green], TTL 250–25,000 (green) corresponds to TTL of ≥ 25,000 copies/µL, Ki67 > 20% or copies/µL [orange] and TTL > 25,000 copies/µL [red], b unadjusted Miller–Payne grade of 3 or 4, Score 3 (orange) corresponds to a TTL (solid lines) or adjusted by Ki67 before NST, Miller–Payne grading of ≥ 25,000 copies/µL, Ki67 > 20% or Miller–Payne grade of 1 or 2, and cN status (dotted lines) by TTL (TTL ≤ 25,000 copies/µL [green] and Score 4 (red) corresponds to a TTL of ≥ 25,000 copies/µL and a and TTL > 25,000 copies/µL [red]), and c according to the DFS prog- Ki67 of > 20%. DFS disease-free survival, NST neoadjuvant systemic nostic score, where Score 1 (blue) corresponds to a TTL of < 25,000 therapy, TTL total tumor load and including only those with clinically negative ALNs TTL (log TTL) and tumor size (non-binary) as predictive [19, 23–25]. Only 62% of those in our study had clinically variables [25]. negative ALN after NST. Since our patients were treated The amount of residual disease, especially in lymph with NST, they were more likely to have more aggressive nodes, after NST remains an important prognostic factor or advanced disease than the patients in these prior stud- [4]. In our study, high TTL (> 25,000 copies/µL) after NST ies. There were, however, some similarities with our patient increased the risk of disease recurrence threefold compared population, including that all or most patients were cT1–3, with low TTL after NST. We found no significant differ- most patients did not have lymphovascular infiltration, and ence in DFS prognosis between patients with a TTL of < 250 all but one study [23] included a majority of patients with copies/µL versus ≥ 250–25,000 copies/µL, suggesting that invasive ductal carcinoma. small metastases (> 250 and < 25,000 copies/µL) detected The nomogram for prediction of non-SLN metastasis by OSNA have similar prognostic value to negative nodes, incorporated three parameters: TTL, tumor size, and a two- i.e. their clinical outcome is no worse than a patient with component histopathologic variable (HER2-negative tumor negative nodes. A difference in DFS prognosis was found at plus Ki67 of ≤ 20%). TTL and tumor size are parameters a cut-off of 25,000 copies/µL, where DFS was significantly common to nomograms predicting non-SLN metastasis shorter among patients with TTL > 25,000 copies/µL than developed by other investigators [24–26]. For example, patients with TTL ≤ 25,000 copies/µL. We propose that this Shimazu and colleagues’ intraoperative nomogram included was related to prior administration of NST: a lower TTL 13 1384 Clinical and Translational Oncology (2021) 23:1377–1385 would be expected in those in whom treatment was effec- after NST. Non-SLN axillary involvement could be pre- tive, indicating a greater likelihood of longer DFS (even if dicted using a nomogram. TTL also had prognostic value, some macrometastases were found), whereas a high TTL allowing classification of patients into different risk grades after NST (> 25,0000 copies/µL) indicated a poor response of disease recurrence and probability of DFS. Although pro- to treatment and these patients would, therefore, be at greater spective studies are required to confirm our findings, our risk of poor DFS. This difference in prognosis remained, study supports the use of the intraoperative OSNA assay in even after adjusting for histopathologic grade, cN status and breast cancer patients treated with NST. Ki67 score. Similar results for this TTL cut-off point have previously been reported in studies in patients that did not Supplementary Information The online version contains supplemen- tary material available at https:​ //doi.org/10.1007/s12094​ -020-02530-​ 4. undergo prior NST [16, 17]. In one of these studies, TTL was prognostic of 5-year DFS at a cut-off of > 25,000 copies/ Acknowledgements The authors would like to thank the Sociedad µL (p = 0.041), and had been determined using intraopera- Española de Senología y Patología Mamaria (SESPM) and Sociedad tive OSNA assay of SLNs [17]. Española de Anatomía Patológica (SEAP) for fostering this study. The We developed a prognostic scoring tool for DFS, which authors also thank Tracy Harrison of Springer Healthcare Communica- tions who provided technical and native-English editing of the manu- included TTL, the Ki67 score before NST, and treatment script prior to submission. This assistance was funded by SYSMEX, response in the primary tumor (Miller–Payne grade). This España S.L. score includes TTL determined from molecular analysis of the SLN, and conventional prognostic markers, and could Author contributions Each author gathered and processed the data of be considered as valuable as other scores of residual dis- his corresponding site. Begoña Vieites and Vicente Peg were the major contributors in writing the manuscript. All authors read and approved ease (for example, the Residual Cancer Burden [21]). The the final manuscript. value of increased accuracy using molecular analysis of axil- lary node status has been previously demonstrated by many Funding This study was supported by a grant from Sysmex España authors [6–12], and is further exemplified in our study by S.L. The sponsor coordinated data collection from study centers, and the observation that the proportion of patients with nodal funded the statistical analysis and medical writing assistance. involvement prior to NST was lower than that as determined Data availability The datasets generated and/or analyzed during the after NST by OSNA assay (24.0% vs. 30.3%). current study are not publicly available due to confidentiality reasons The potential clinical value of predicting non-SLN but are available from the corresponding author on reasonable request. involvement using TTL and a nomogram among breast cancer patients post-NST is to spare the patient undergoing Compliance with ethical standards ALND when there is a low probability of ALN metastases. While the ACOSOG Z001 trial demonstrated that ALND Conflict of interests The authors declare that they have no competing can be avoided in a select group of SLN-positive patients interests. [27], patients who have received NST such as those in our Ethics approval The study was conducted in accordance with the Dec- study do not meet ACOSOG Z001 criteria, and indeed appli- laration of Helsinki and guidelines for Good Clinical Practice, after cation of these criteria seems to exclude the majority of obtaining PEIBA Ethics Committee (Portal de Ética de la Investigación SLN-positive patients [25]. Since an OSNA assay to deter- Biomédica de Andalucía) approval. mine TTL and thus non-SLN involvement can be performed Informed consent The ethic committee confirmed that there was no intraoperatively [25, 26], it can also offer the advantage of need of informed consent for this study. clinical decision-making during primary surgery, avoiding a second surgery for ALND if required. Open Access This article is licensed under a Creative Commons Attri- Study limitations attributable to the retrospective nature bution 4.0 International License, which permits use, sharing, adapta- of the design included a lack of standardized NST and lack tion, distribution and reproduction in any medium or format, as long of a comparator control group. Also, molecular analysis by as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes OSNA may not be as commonly available worldwide as it were made. The images or other third party material in this article are is in Europe and Japan. Finally, it may be argued by some included in the article’s Creative Commons licence, unless indicated that the use of whole nodes in the OSNA assay, precluding otherwise in a credit line to the material. If material is not included in histologic analysis of the node, is a limitation. But multiple the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will studies have shown that molecular assay is as accurate as need to obtain permission directly from the copyright holder. To view a conventional pathologic assessment, including in the post- copy of this licence, visit http://creat​iveco​mmons​.org/licen​ses/by/4.0/. NST setting [28]. In conclusion, this is the first fully published study to establish the predictive and prognostic value of TTL derived from molecular analysis of the SLN in breast cancer patients 13 Clinical and Translational Oncology (2021) 23:1377–1385 1385 References 16. Peg V, Sansano I, Vieites B, et al. Role of total tumour load of sentinel lymph node on survival in early breast cancer patients. Breast. 2017;33:8–13. 1. World Cancer Research Fund and American Institute for Cancer 17. Piñero-Madrona A, Ripoll-Orts F, Sánchez-Méndez JI, et al. Research. Worldwide cancer data: Global cancer statistics from External validation of a prognostic model based on total tumor the most common cancers. WCRF.org. 2020. https​://www.wcrf. load of sentinel lymph node for early breast cancer patients. Breast org/dietan​ dcanc​ er/cancer​ -trends​ /worldw ​ ide-cancer​ -data. Accessed Cancer Res Treat. 2020;181(2):339–45. 29 June 2020. 18. Ogston KN, Miller ID, Payne S, et al. A new histological grad- 2. Untch M, Konecny GE, Paepke S, von Minckwitz G. Current ing system to assess response of breast cancers to primary and future role of neoadjuvant therapy for breast cancer. Breast. chemotherapy: prognostic significance and survival. Breast. 2014;23(5):526–37. 2003;12(5):320–7. 3. van Nes JG, Putter H, Julien JP, et al. Preoperative chemotherapy 19. Peg V, Espinosa-Bravo M, Vieites B, et al. Intraoperative molecu- is safe in early breast cancer, even after 10 years of follow-up; lar analysis of total tumor load in sentinel lymph node: a new clinical and translational results from the EORTC trial 10902. predictor of axillary status in early breast cancer patients. Breast Breast Cancer Res Treat. 2009;115(1):101–13. Cancer Res Treat. 2013;139(1):87–93. 4. Cardoso F, Kyriakides S, Ohno S, et al. Early breast cancer: 20. Mahul B Amin M, Edge SB, and the American Joint Commit- ESMO Clinical Practice Guidelines for diagnosis, treatment and tee on Cancer. AJCC Cancer Staging Manual. 8th ed. Springer, follow-updagger. Ann Oncol. 2019;30(8):1194–220. Switzerland, 2017. 5. Hunter-Smith AE, Rayter Z. One-step nucleic acid amplifica- 21. Symmans WF, Peintinger F, Hatzis C, et al. Measurement of tion: the possible value in assessing sentinel lymph node metas- residual breast cancer burden to predict survival after neoadjuvant tasis during mastectomy. Breast Cancer (Dove Med Press). chemotherapy. J Clin Oncol. 2007;25(28):4414–22. 2018;10:13–21. 22. Rubio IT, Espinosa-Bravo M, Rodrigo M, et al. Nomogram 6. Bernet L, Cano R, Martinez M, et al. Diagnosis of the sentinel including the total tumoral load in the sentinel nodes assessed by lymph node in breast cancer: a reproducible molecular method: a one-step nucleic acid amplification as a new factor for predict- multicentric Spanish study. Histopathology. 2011;58(6):863–9. ing nonsentinel lymph node metastasis in breast cancer patients. 7. Feldman S, Krishnamurthy S, Gillanders W, et al. A novel auto- Breast Cancer Res Treat. 2014;147(2):371–80. mated assay for the rapid identification of metastatic breast carci- 23. Fung V, Kohlhardt S, Vergani P, Zardin GJ, Williams NR. Intra- noma in sentinel lymph nodes. Cancer. 2011;117(12):2599–607. operative prediction of the two axillary lymph node macrome- 8. Schem C, Maass N, Bauerschlag DO, et al. One-step nucleic acid tastases threshold in patients with breast cancer using a one-step amplification-a molecular method for the detection of lymph node nucleic acid cytokeratin-19 amplification assay. Mol Clin Oncol. metastases in breast cancer patients; results of the German study 2017;7(5):755–62. group. Virchows Arch. 2009;454(2):203–10. 24. Sa-Nguanraksa D, OC E, Kulprom A et al. Nomogram to predict 9. Shi F, Liang Z, Zhang Q, Wang C, Liu X. The performance of non-sentinel lymph node status using total tumor load determined one-step nucleic acid amplification assay for intraoperative detec- by one-step nucleic acid amplification: first report from Thailand. tion of sentinel lymph node macrometastasis in breast cancer: An Breast Cancer. 2019;26(4):471–7. updated meta-analysis. Breast. 2018;39:39–45. 25. Shimazu K, Sato N, Ogiya A, et al. Intraoperative nomograms, 10. Tsujimoto M, Nakabayashi K, Yoshidome K, et al. One-step based on one-step nucleic acid amplification, for prediction of nucleic acid amplification for intraoperative detection of lymph non-sentinel node metastasis and four or more axillary node node metastasis in breast cancer patients. Clin Cancer Res. metastases in breast cancer patients with sentinel node metastasis. 2007;13(16):4807–16. Ann Surg Oncol. 2018;25(9):2603–11. 11. Visser M, Jiwa M, Horstman A, et al. Intra-operative rapid diag- 26. Sun X, Zhang Y, Wu S, Fu L, Yun JP, Wang YS. Intraoperative nostic method based on CK19 mRNA expression for the detec- prediction of non-sentinel lymph node metastasis based on the tion of lymph node metastases in breast cancer. Int J Cancer. molecular assay in breast cancer patients. Cancer Manag Res. 2008;122(11):2562–7. 2019;11:9715–23. 12. Osako T, Tsuda H, Horii R, et al. Molecular detection of lymph 27. Giuliano AE, Hunt KK, Ballman KV, et al. Axillary dissection node metastasis in breast cancer patients treated with preop- vs no axillary dissection in women with invasive breast cancer erative systemic chemotherapy: a prospective multicentre trial and sentinel node metastasis: a randomized clinical trial. JAMA. using the one-step nucleic acid amplification assay. Br J Cancer. 2011;305(6):569–75. 2013;109(6):1693–8. 28. Espinosa-Bravo M, Navarro-Cecilia J, Ramos Boyero M, et al. 13. Vieites B, Lopez-Garcia MA, Castilla C, et al. CK19 expression Intraoperative assessment of sentinel lymph node by one-step in breast tumours and lymph node metastasis after neoadjuvant nucleic acid amplification in breast cancer patients after neoadju- therapy. Histopathology. 2016;69(2):239–49. vant treatment reduces the need for a second surgery for axillary 14. Navarro-Cecilia J, Duenas-Rodriguez B, Luque-Lopez C, et al. lymph node dissection. Breast. 2017;31:40–5. Intraoperative sentinel node biopsy by one-step nucleic acid amplification (OSNA) avoids axillary lymphadenectomy in Publisher’s Note Springer Nature remains neutral with regard to women with breast cancer treated with neoadjuvant chemotherapy. jurisdictional claims in published maps and institutional affiliations. Eur J Surg Oncol. 2013;39(8):873–9. 15. Parada D, Pena KB, Riu FF, Aguilar AE, Cohan S. Intraoperative molecular analysis of sentinel lymph nodes following neoadjuvant chemotherapy in patients with clinical node negative breast can- cer: An institutional study. Mol Clin Oncol. 2016;5(5):507–10. 13