Axillary Surgery in Breast Cancer: Evidence-Based De-escalation Across Upfront and Post-Neoadjuvant Settings.

INTRODUCTION
Axillary lymph node status has long been a major prognostic factor in breast cancer, guiding both adjuvant therapy and locoregional management. Axillary lymph node dissection (ALND), standardized in Halsted’s radical mastectomy in the 1890s, was based on the belief of stepwise tumor spread from the breast to the axilla and beyond. Removal of level I–III nodes achieved local control and provided the most accurate staging before the era of sentinel lymph node biopsy (SLNB) and genomic profiling. Despite complications such as lymphedema, shoulder stiffness, and chronic pain, ALND remained the gold standard for decades because it ensured both staging accuracy and regional disease control [1].
The introduction of SLNB in the 1990s marked a paradigm shift in axillary management. By identifying the first draining lymph node from the breast, SLNB achieved staging accuracy comparable to ALND while greatly reducing morbidity. Within a decade, it became the global standard for clinically node-negative patients, significantly lowering rates of lymphedema and functional impairment. This milestone transformed surgical practice and initiated the modern concept of axillary de-escalation, i.e., shifting the goal from radical clearance to selective, evidence-based intervention according to nodal burden [23].
Subsequent randomized controlled trials further questioned the need for completion ALND even in patients with sentinel node metastasis. The eligible population gradually expanded according to nodal burden. IBCSG 23-01 and AATRM enrolled patients with one or more sentinel nodes containing only micrometastases (≤ 2 mm) and showed that omission of ALND did not compromise survival or locoregional control. ACOSOG Z0011 and AMAROS included patients with one to two positive sentinel nodes, encompassing both micro- and macrometastases, and confirmed the safety of omitting ALND when systemic therapy and radiotherapy (RT) were appropriately administered. More recently, SENOMAC and SINODAR-ONE extended this evidence to patients with one or two macrometastases (> 2 mm), demonstrating oncologic safety in well-selected cases. Collectively, these trials established a continuum of axillary de-escalation—from micrometastatic to macrometastatic disease—defining the modern framework of selective axillary management [4567891011].
Parallel to these advances in upfront surgery, a second field of axillary optimization has evolved in the neoadjuvant chemotherapy (NAC) setting. For patients initially presenting with cN0 disease, SLNB after chemotherapy is now well established, showing identification rates (IRs) above 90% and acceptable false-negative rates (FNRs). In patients initially node-positive, refinements such as dual tracer mapping, retrieval of three or more sentinel nodes, and targeted axillary dissection (TAD) have improved staging accuracy and reduced the need for routine ALND [12]. Moreover, recent data, including the OPBC-05/ICARO study, indicate that even when residual isolated tumor cells (ITCs) or micrometastases remain after NAC, completion ALND may be safely omitted in carefully selected patients [1314].
Looking ahead, several ongoing phase III trials are extending the boundaries of axillary de-escalation. ASICS, EUBREAST-01, and ASLAN are testing whether SLNB can be safely omitted in excellent responders to neoadjuvant therapy, particularly in human epidermal growth factor receptor 2-positive (HER2+) and triple-negative disease with high rates of pathologic complete response (pCR). In the upfront setting, SOUND, INSEMA, and NAUTILUS are evaluating the omission of SLNB in patients with negative axillary imaging. Collectively, these studies represent a growing movement toward minimizing or even eliminating axillary surgery in well-selected populations [151617].
In the contemporary era, axillary surgery has shifted from maximal clearance toward optimization. The current challenge is to balance oncologic safety with reduced morbidity while integrating systemic therapy response and tumor biology into decision-making. The future of axillary management lies in individualized strategies guided by high-quality clinical evidence, advanced imaging, molecular profiling, and multidisciplinary collaboration.

DE-ESCALATION OF UPFRONT AXILLARY SURGERY IN CN0 BREAST CANCER

Transition from routine ALND to a selective SLNB approach
In cN0 breast cancer, the optimization of axillary management in the upfront surgery setting began with investigations into whether ALND could be safely omitted in patients with only micrometastatic disease, and this concept was progressively extrapolated and expanded to groups with a higher metastatic burden. This stepwise process was established by a series of randomized clinical trials conducted from the early 2000s through the mid-2020s.
The earliest high-level evidence came from the trials on micrometastasis. The IBCSG 23-01 trial, which enrolled patients between 2001 and 2010, was a phase III randomized study of women with tumors ≤ 5 cm and sentinel lymph nodes (SLNs) harboring only micrometastases (≤ 2 mm) without extracapsular extension [618]. All patients were cN0 prior to surgery and were designed to receive standard adjuvant systemic and radiation therapy. The trial demonstrated non-inferiority of omitting ALND in terms of 5-year disease-free and overall survival (OS), and axillary recurrence was reported at less than 1%. During the same period, the AATRM trial (2001–2008) in Northern Spain included patients with tumors ≤ 3.5 cm, cN0 status, and micrometastases in the SLN, comparing ALND to clinical follow-up [11]. With approximately five years of follow-up, no significant differences in disease-related outcomes were observed. In both studies, additional non-SLN metastases were identified in about 13% of the ALND arms, but this additional pathological information did not translate into improved long-term clinical outcomes, whereas the risks of lymphedema and neurological complications associated with ALND became more evident. Collectively, these trials conducted in the early 2000s established the clinical turning point that ALND omission is safe in low-burden disease limited to micrometastasis.
The next stage involved the mixed-burden trials, which allowed enrollment of patients with one or two positive SLNs regardless of the size of nodal metastasis, thereby including both micrometastatic and macrometastatic disease. The ACOSOG Z0011 trial, conducted in North America between 1999 and 2004, enrolled patients with clinical T stage (cT1–2) invasive breast cancer, no palpable axillary lymphadenopathy, and one to two positive SLNs. All participants were scheduled to undergo breast-conserving surgery with tangential whole-breast irradiation and systemic therapy. After 10 years of follow-up, OS and disease-free survival (DFS) were equivalent between the ALND and SLNB-alone groups, and regional recurrence remained very low, even though additional non-SLN metastases were identified in approximately 27% of the ALND arm [28]. The AMAROS trial, conducted in Europe between 2001 and 2010, randomized cT1–2N0 patients with positive SLNs to ALND or axillary RT (ART) [10]. Approximately 82% of patients underwent breast-conserving surgery, and 18% underwent mastectomy; about 40% of patients had micrometastases or ITCs. At 10 years, axillary recurrence, locoregional recurrence, DFS, and OS were equivalent between the two groups, but the incidence of lymphedema was significantly higher in the ALND arm compared with ART. Together, these mixed-burden trials established that completion ALND is no longer essential when systemic therapy and RT are appropriately administered, even in patients with one or two positive SLNs.
The most recent body of evidence comes from the macrometastasis trials, which deliberately restricted enrollment to patients with one or two SLNs containing macrometastases (≥ 2 mm), thereby testing the safety of omitting ALND in higher-burden disease. The SENOMAC trial, conducted across Nordic countries between 2015 and 2021, enrolled cN0 patients with positive SLNs, including those with T1–T3 tumors, extracapsular extension, and mastectomy, thus broadening eligibility criteria [5]. With a median follow-up of approximately 47 months, no differences in breast cancer–specific survival or DFS were observed between the ALND and omission arms. Notably, among patients who underwent ALND, about one-third harbored additional non-SLN metastases, and approximately 13% were upstaged to pN2 or pN3 disease, yet this pathological information did not translate into superior clinical outcomes. The Italian SINODAR-ONE trial, enrolling patients between 2015 and 2020, included women with cT2 (≤ 5 cm) cN0 breast cancer and one to two SLNs with macrometastases [9]. Breast-conserving surgery was performed in approximately 75% and mastectomy in 23% of cases. At three years, OS and recurrence outcomes were equivalent between arms despite 44% of patients in the ALND arm had additional non-SLN macrometastases. The ongoing POSNOC trial, which began enrolling in 2014 across the UK, Australia, and New Zealand, is designed as a non-inferiority trial. It includes patients with unifocal or multifocal invasive breast cancers ≤ 5 cm and one or two SLNs with macrometastases, regardless of extranodal extension, and randomizes them to adjuvant therapy alone vs. adjuvant therapy plus axillary treatment (ALND or ART) [19]. The primary endpoint is 5-year axillary recurrence. This trial is prospectively testing the hypothesis that even in higher-burden disease, ALND may not be universally required when high-quality systemic and radiation therapy are delivered.
Taken together, this continuum of evidence delivers a consistent message both in chronological sequence and in the precision of patient selection. In the early 2000s, ALND omission was first validated in low-burden populations limited to micrometastasis. From the late 2000s through the 2010s, the principle expanded to mixed-burden cohorts including both micrometastases and macrometastases, demonstrating safety in the context of breast-conserving surgery, whole-breast irradiation, and appropriate systemic therapy. Since the mid-2010s, randomized trials intentionally restricted to patients with macrometastases have further shown that one or two positive SLNs do not, in themselves, justify routine completion ALND. Across these studies, a consistent theme emerges: although a proportion of patients in the ALND arms had additional nodal disease detected, this did not translate into meaningful improvements in survival or recurrence outcomes, whereas the morbidity burden of ALND such as lymphedema and functional impairment remains significant. Therefore, in the upfront surgical management of cN0 breast cancer, the goal of axillary surgery is no longer maximal clearance but rather optimization, ensuring that treatment is proportional to disease burden and therapeutic context. With the integration of modern systemic therapy and RT, and with precise patient selection, omission of ALND, and in some settings, even omission of SLNB has emerged as a safe and rational option (Table 1) [58910111819].

Omission of SLNB in carefully selected cN0 breast cancer
The contemporary question in cN0 early breast cancer is whether surgical axillary staging can be omitted altogether in carefully selected patients—specifically those with small tumors and negative axillary imaging—without compromising oncologic outcomes. Two large, randomized trials, SOUND and INSEMA, provide complementary high-level evidence that observation (no axillary surgery) is non-inferior to SLNB in this setting, with very low axillary recurrence and measurable morbidity benefits.
SOUND was designed in Italy as a pragmatic, multicenter, randomized noninferiority trial enrolling women with small invasive breast cancers and negative preoperative axillary ultrasonography (AUS) [20]. Key enrollment features were: 1) cN0 axilla, 2) tumor size up to 2 cm (T1), and 3) mandatory negative AUS with predefined criteria; patients were randomized to no axillary surgery vs. SLNB. The primary endpoint was 5-year distant DFS (DDFS). The trial enrolled during the 2010s (trial registration NCT02167490) and reported its main outcomes in JAMA Oncology in November 2023, demonstrating noninferiority of observation to SLNB and a very low cumulative axillary recurrence (0.4% at 5 years in both arms) and 5-year DDFS 97.7% vs. 98.0% (hazard ratio [HR], 0.84; 90% confidence interval [CI], 0.45–1.54) despite a 13.7% node-positive rate among those who underwent SLNB. These data suggest that, where negative AUS is present and pathologic node status would not alter systemic or radiation decisions, SLNB confers limited additional oncologic value while still exposing patients to surgical morbidity.
Being conducted primarily in Germany and Austria, INSEMA tested SLNB omission in a broader cN0 population than SOUND. It enrolled T1–T2 (≤ 5 cm) invasive cancers scheduled for breast-conserving surgery, randomizing patients to no axillary surgery vs. SLNB (1:4 allocation). The per-protocol analysis (n = 4,858) with median follow-up 73.6 months showed noninferiority of omission for 5-year invasive DFS (IDFS) (91.9% vs. 91.7%; HR, 0.91; 95% CI, 0.73–1.14), with axillary recurrence 1.0% vs. 0.3% and less lymphedema, better arm mobility, and less pain in the omission arm. Compared with SOUND, INSEMA permitted larger tumors (up to 5 cm), required breast-conserving surgery, and used IDFS (not DDFS) as the primary endpoint, extending generalizability to a broader, predominantly hormone receptor-positive (HR+)/human epidermal growth factor receptor 2-negative (HER2−) cohort while reinforcing the morbidity advantages of omission [21].
NAUTILUS is a Korean multicenter randomized trial explicitly modeled to test SLNB omission in the AUS-negative cT1–T2N0 population undergoing breast-conserving surgery with whole-breast irradiation, which is conceptually aligned with SOUND/INSEMA but tailored to Asian epidemiology [22]. Enrollment began in September 2020 with a planned sample of 1,734 patients and primary endpoint of 5-year IDFS; all patients receive standardized whole-breast RT. NAUTILUS mandates predefined AUS criteria, allows US-guided core needle biopsy/FNA for minimally suspicious nodes before randomization, and blinds patients to their assigned arm until surgery—methodological refinements intended to reduce misclassification and preference bias. Like INSEMA, NAUTILUS uses IDFS (not DDFS) as the primary endpoint, but, akin to SOUND, it anchors selection on negative standardized AUS; importantly, it is the only large randomized controlled trial focused on an Asian cohort (younger, more often premenopausal), addressing external validity across populations. Protocol details were published in 2022 and registered at ClinicalTrials.gov (NCT04303715). Together, SOUND, INSEMA, and NAUTILUS outline a coherent strategy, ‘negative AUS plus planned breast-conserving therapy’ define a subgroup in which omitting SLNB is oncologically safe and reduces surgery-related morbidity. NAUTILUS adds rigorous imaging criteria and geographic/biologic representation likely to influence guideline adoption in Asia.
In summary, SOUND confirmed the safety of observation over SLNB in T1 tumors with negative AUS, INSEMA extended this to T1–T2 cases with quality-of-life benefit, and NAUTILUS applies the same concept using standardized AUS and mandatory WBI in an Asian cohort. In addition to these landmark trials, other ongoing studies such as SOAPET and BOOG 13-08 are expected to further validate the omission strategy in cN0 patients with negative imaging, including positron emission tomography-computed tomography-based and Dutch multicenter protocols that will refine imaging-guided selection criteria (Table 2) [202122232425].

AXILLARY MANAGEMENT IN THE POST-NEOADJUVANT SETTING

SLNB after NAC in initial cN0 disease
For patients who are cN0 at diagnosis and receive NAC, performing SLNB after NAC achieves test performance that is close to upfront surgery. The most focused meta-analysis confined to this population (16 studies; n = 1,456) reported a pooled IR of 96% (95% CI, 95–97) and a FNR of 6% (95% CI, 3–8), concluding that SLNB after NAC is “technically feasible and accurate enough” for axillary staging in initially cN0 disease. These results were consistent across mapping techniques and histopathologic protocols, with high sensitivity and negative predictive value as well [7]. Contemporary practice reviews from major societies have therefore placed SLNB after NAC as an appropriate standard for staging the axilla in cN0 patients, emphasizing the practical advantage of consolidating breast and axillary surgery into a single operation without compromising regional control when SLNs are negative and adjuvant therapy is delivered per current standards.

SLNB after NAC in initial clinically node-positive disease
In contrast, for patients who are biopsy-proven node-positive (cN+) at baseline and convert to clinically node-negative after neoadjuvant chemotherapy (ycN0) after NAC, SLNB performance is inferior to the cN0 setting, though it can be optimized with technique. In the landmark ACOSOG Z1071 trial [26], among women presenting with cN1 disease who underwent SLNB after NAC followed by ALND, the FNR was 12.6% (in those with ≥ 2 SLNs examined), prompting calls for technical refinements to improve sensitivity.
Prospective trials consistently show that retrieving more sentinel nodes and using dual-tracer mapping improve accuracy. In SENTINA [27], when SLNB was performed after NAC in patients who converted from cN+ to ycN0, the FNR was 24.3% if only one SLN was removed and 18.5% when two were removed—illustrating why most protocols now target ≥ 3 SLNs to keep FNR within acceptable bounds. Meta-analyses focused on initially node-positive cohorts likewise report IR ≈ 88%–91% and FNR ≈ 13%–14%, with dual mapping and greater SLN yield associated with better performance (Table 3) [726272829].
Technique refinements that ensure removal of the biopsy-proven positive (clipped) node further reduce FNR. A prespecified analysis from ACOSOG Z1071 demonstrated that when the clipped node was identified within the SLN specimen, the FNR fell to 6.8%. These observations underpin the evolution toward TAD, which combines SLNB with focused retrieval of the clipped node and has repeatedly shown single-digit FNRs in prospective series and systematic reviews (Table 4) [26303132].
Finally, outcomes data support that these technical thresholds matter clinically. In a large single-center cohort of patients initially cN1 who became ycN0 after NAC, proceeding with SLNB alone was associated with low nodal recurrence when ≥ 3 SLNs were retrieved and all were negative under dual-tracer mapping, reinforcing the operational rule set (dual tracer, ≥ 3 SLNs, and—where feasible—clipped-node retrieval) for safely avoiding routine ALND in this setting [26273033].

Omission of completion ALND in residual nodal disease after NAC
Two evidence-based omission pathways are now shaping post-NAC axillary management. The first concerns forgoing completion ALND when only low-volume residual nodal disease remains after NAC (e.g., ITCs or micrometastases). The second moves further upstream, testing omission of any axillary surgery (including SLNB) in rigorously selected excellent responders.

Low-volume residual nodal disease after NAC: evidence for selective omission of completion ALND
Recent multicenter analyses have questioned whether residual ITCs or micrometastases after NAC necessitate completion ALND.
The OPBC-05/ICARO study was a global retrospective registry including 583 patients with pathologic isolated tumor cells only after NAC (ypN0(i+)) disease after NAC. Under contemporary systemic therapy and regional irradiation, 3-year axillary recurrence was only ≈ 2%, with no difference between ALND and non-ALND groups, indicating that routine dissection may be unnecessary in this low-burden setting. However, no prospective data are yet available for this population [13].
The OPBC-07/microNAC consortium extended these findings to > 2,000 patients with residual micrometastatic disease (0.2–2 mm). The 5-year axillary recurrence remained ≈ 2%, and overall DFS/OS were comparable regardless of ALND. Notably, in the triple-negative breast cancer (TNBC) subgroup, omission of ALND was associated with significantly worse 5-year DFS (~p = 0.03) despite low axillary failure (< 2%) [14].
Together, these retrospective OPBC datasets define the current evidence base supporting selective, rather than universal, ALND after NAC in patients with low-volume residual nodal disease, while underscoring the need for prospective validation (Table 5) [1314].

Excellent responders to NAC: toward complete omission of SLNB itself
A forward-leaning strategy is to eliminate axillary surgery in initially cN0 patients who, after NAC, achieve an excellent radiologic response (and often breast pathologic complete response [bpCR]), particularly in HER2+ and triple-negative subtypes that display the highest pCR rates. Three coordinated initiatives are emblematic: EUBREAST-01, ASICS, and ASLAN. All share the core premise that, in such enriched cohorts, nodal surgery adds little oncologic value yet carries morbidity; endpoints center on invasive disease-free outcomes and carefully monitored regional control under standardized adjuvant RT.
EUBREAST-01 (NCT04101851) is the first multinational program to formalize SLNB omission after NAC in biologically favorable subtypes. Conducted under the auspices of the German Breast Group, the trial enrolled patients with cT1–3, cN0, HER2+ or TNBC disease who achieved radiologic complete response (CR) on magnetic resonance imaging (MRI) and pathologic CR in the breast after NAC. All axillary surgery was omitted, while breast-conserving surgery was performed per institutional standards. RT was standardized through national protocols emphasizing regional coverage. The primary endpoint is 3-year axillary recurrence-free survival (ARFS), with secondary endpoints including locoregional recurrence, any invasive recurrence, and OS. Recruitment was completed in December 2024 (≈ 267 analyzable of 365 screened), and data maturation is ongoing. EUBREAST-01 established the initial feasibility and safety framework for complete omission of axillary surgery in imaging-confirmed pCR cohorts [16].
ASICS (NCT04225858) study, coordinated in the Netherlands, complements EUBREAST-01 by broadening eligibility and incorporating patient-reported outcomes (PROs). This prospective single-arm trial also focuses on HER2+ and TNBC patients achieving radiologic CR on MRI after NAC, allowing both breast-conserving surgery and mastectomy. Axillary surgery is omitted in all qualifying patients, and RT protocols are standardized but recorded for institutional variation. The primary endpoint is ARFS, and key secondary outcomes include locoregional recurrence, invasive recurrence, morbidity, and quality of life. Recruitment is ongoing (2020–2025), and early feasibility data suggest consistency with EUBREAST-01 findings. Together, these two European trials will provide foundational evidence that axillary surgery may be dispensable in MRI-defined complete responders, given optimized systemic and RT backbones [15].
ASLAN (selective Avoidance of Sentinel Lymph node biopsy After Neoadjuvant chemotherapy) trial (KBCSG-28; NCT04993625) represents the Asian counterpart, designed to translate this concept into practice within a distinct demographic and healthcare environment [17]. It is an academically sponsored, prospective, multicenter, single-arm study led by Korean Breast Cancer Society. ASLAN specifically targets HER2+ and TNBC patients who exhibit excellent radiologic response in both the breast and the axilla following NAC, using a rigorously defined dual-compartment clearance criterion based on MRI and ultrasonography. Under the protocol, SLNB is completely omitted when pCR in lumpectomy is confirmed, and all patients proceed to the standardized whole-breast and regional RT. Centralized RT plan review and a prospectively embedded PRO program are integral components of the design, reflecting its focus on both oncologic and quality-of-life outcomes. As of July 2025, 254 patients had been screened, of whom 182 were enrolled (two exclusions due to RT refusal or follow-up loss). Enrollment closed in December 2023, and analysis milestones are pre-specified: a mid-database lock in December 2025, interim presentation in April 2026 once approximately half of the cohort reaches 3-year follow-up, and final analysis in 2029 after 5-year follow-up completion is anticipated. Compared with EUBREAST-01 and ASICS, ASLAN distinguishes itself through 1) an Asian, predominantly premenopausal population; 2) explicit dual-compartment imaging clearance and bpCR in lumpectomy as the prerequisite for omission of SLNB; 3) standardized RT and embedded PRO assessment; and 4) a transparent dissemination timeline linked to predefined data locks. These characteristics make ASLAN not only a scientific validation effort but also an implementation-oriented trial, designed to provide real-world guidance for regions where HER2-targeted agents and immune-checkpoint inhibitors are integrated into NAC regimens.
Collectively, EUBREAST-01, ASICS, and ASLAN illustrate a convergent global movement toward eliminating axillary surgery in highly selected excellent responders after NAC. As data matures, these coordinated programs are expected to define the oncologic boundaries of axillary omission and to inform future international guidelines on patient-centered, biology-adapted management of the axilla (Table 6) [1516173435].

SPECIAL CONSIDERATIONS IN AXILLARY DE-ESCALATION

Axillary de-escalation in patients undergoing mastectomy
The Dutch nationwide BOOG 2013-07 registry analyzed 1,090 mastectomy patients with limited SLN metastases (pN1mi–pN1a, 1–3 nodes). Axillary management varied widely—no further treatment (20%), ALND (40%), regional nodal irradiation (RNI) (30%), or ALND + RNI (10%)—yet 5-year regional recurrence remained low (1.3%) without intergroup difference. Selection bias and non-standardized RT policies, however, limit generalizability [24]. Complementary evidence from the SENOMIC cohort (including mastectomy patients) similarly showed very low axillary recurrence (~3%) after omission of ALND in SLN-positive cases [36]. The IBCSG 23-01 trial, though predominantly breast-conserving, supported omission for micrometastases with consistent direction of effect [18]. Recent retrospective series further reported no survival disadvantage with ALND omission in carefully selected mastectomy patients [3738].
Taken together, available data suggests omission may be oncologically safe when modern systemic therapy and tailored post-mastectomy RT are applied, but randomized, mastectomy-exclusive trials remain lacking.

Systemic therapy decision-making without surgical upstaging
The MonarchE program demonstrated that adjuvant Abemaciclib plus endocrine therapy significantly improves IDFS in patients with HR+/HER2−, node-positive, high-risk early breast cancer [39]. In October 2025, the Annals of Oncology publication reported the primary OS outcomes at a median follow-up of 6.3 years, showing statistically significant OS benefit (HR, 0.84; 95% CI, 0.72–0.98, p = 0.027) together with sustained improvements in IDFS and distant relapse-free survival [40]. Trial eligibility stratified risk by node count (≥ 4 positive nodes) or 1–3 nodes with high-risk features (tumor ≥ 5 cm, grade 3, or high Ki-67), but did not require completion ALND to enumerate additional nodes beyond SLNB. These data, and contemporary expert reviews, support the premise that systemic therapy should be guided by available pathology from SLNB and tumor biology, and that ALND should not be performed solely to upstage node counts for systemic-therapy access.
On the other hand, risk-prediction tools may further reduce reliance on surgical confirmation of extensive axillary burden. Classic non-SLN nomograms (MSKCC; MD Anderson) estimate the probability of additional positive nodes after a positive SLN and have undergone multiple external validations with moderate discrimination—useful for counseling and RT planning, but not a substitute for pathology when treatment hinges on nodal status [4142].

Surgery vs. RT for regional control
Randomized evidence indicates that ART (RNI) can match the regional control of surgery with less morbidity in SLN-positive disease. In AMAROS (EORTC 10981–22023), 10-year results confirmed no differences in axillary recurrence, DFS, or OS between ALND and ART, while lymphedema remained significantly lower with ART (24.5% vs. 11.9% at 5 years) [4]. The single-center OTOASOR trial likewise showed equivalent locoregional control and survival between completion ALND and RNI after a positive SLN, supporting RT as a less morbid alternative for regional control (Table 7) [4581020254344454647].
In the post-neoadjuvant setting, the question becomes whether additional RNI is needed when initially node-positive patients convert to ypN0. The phase III NSABP B-51/RTOG 1304 trial found that adding RNI did not improve invasive breast cancer recurrence-free interval, locoregional control, distant control, DFS, or OS compared with no RNI in this ypN0 population—supporting de-escalation of regional therapy when modern systemic therapy yields nodal pCR [45].

FUTURE DIRECTIONS
The future of axillary management lies not in uniform de-escalation but in precision optimization tailored to each patient’s disease biology and treatment response. In the upfront setting, trials such as SOUND, INSEMA, and NAUTILUS are demonstrating that surgical omission can be safely applied when preoperative imaging reliably excludes nodal involvement. In the neoadjuvant setting, studies including ASICS, EUBREAST-01, and ASLAN are testing omission of axillary surgery in biologically favorable subtypes achieving excellent response. These programs represent a shift from procedural uniformity toward evidence-based individualization, where surgery is refined rather than minimized. Historically, postoperative morbidity was considered an inevitable trade-off for oncologic control; however, the modern paradigm emphasizes maintaining oncologic safety while maximizing quality of life. Future clinical integration of imaging assessment, molecular profiling, and systemic therapy response will further enable selective, biology-driven axillary management that achieves cure through optimization—not escalation.

CONCLUSION
Axillary surgery in breast cancer has evolved from a paradigm of maximal clearance to one of precision and restraint. Accumulated randomized evidence demonstrates that extensive axillary dissection provides little additional oncologic benefit when contemporary systemic therapy and RT are appropriately delivered. Completion ALND can be safely omitted in patients with one or two positive sentinel nodes, and even SLNB itself may be unnecessary in selected patients with negative axillary imaging and favorable tumor biology.
In the neoadjuvant setting, optimized techniques such as dual-tracer mapping, retrieval of three or more sentinel nodes, and TAD allow accurate staging and safe omission of ALND in patients who convert to ycN0. For excellent responders—particularly those with HER2+ or triple-negative subtypes—prospective trials including EUBREAST-01, ASICS, and ASLAN are redefining the boundaries of axillary surgery, potentially permitting complete omission in rigorously selected cases.
Beyond surgery, systemic therapy and RT have assumed pivotal roles in regional disease control. Modern adjuvant indications, exemplified by the MonarchE trial, are increasingly based on biologic and clinical risk rather than surgical node counts, while randomized trials such as AMAROS, OTOASOR, and NSABP B-51 confirm that regional RT can provide equivalent control with reduced morbidity.
The future of axillary care lies in biologically driven, response-adapted strategies that maximize cure while minimizing harm. The guiding principle is no longer “how many nodes to remove,” but rather “how little is enough”—achieving oncologic safety through integration, not escalation.