A total of 1158 articles were identified from PubMed, Embase, and Cochrane Library for title screening. There were 14 non-English and 189 duplicated articles, which were excluded. After screening the titles and abstracts, 260 were reviews, commentaries, or conference abstracts, and 671 were found irrelevant. A total of 24 articles were eligible for full-text review. Among these, the oncological outcome of patients with successful nodal conversion and axillary surgery de-escalation was not clearly described in five studies. Another five articles were excluded for overlapping data published on different occasions [14,15,16,17,18]. As a result, 11 retrospective studies [20,21,22,23,24,25,26,27,28,29,30] and 3 prospective studies [19, 31, 32] matched the inclusion criteria and were included in this meta-analysis (Fig. 1). Detailed information on the reviewed articles is displayed in Table 1.

PRISMA flowchart for the study

The included cohorts consisted of 14 studies published between 2018 and 2023. All studies achieved an overall score of 7 or above on the Newcastle–Ottawa Scale [13] (Table 1). Eleven studies focused on the oncological outcomes of SLNB alone [19,20,21,22,23,24,25,26,27,28,29], while the remaining three articles assessed newer surgical modalities of MARI [30] and TAD [31, 32]. They comprised a total of 4268 patients with node-positive breast cancers before NST.

Majority of the recruited patients had clinical T1-3 (95.0%) and N1 (78.9%) breast cancers (Table 1). Eleven studies reported on the biological subtypes of primary breast tumors [20, 21, 23, 25,26,27,28,29,30,31,32]. Among these, 39.6% were hormone receptor (HR)-positive and human epidermal growth factor receptor 2 (HER2)-negative tumors; 20.9% were HR-positive and HER2-positive; 16.2% were HER2-positive only and 23.0% were triple negative. Five studies only recruited patients who were converted to clinical or pathological node-negative breast cancers after NST [19, 22, 23, 25, 27]. Patients with inflammatory breast cancers [23, 25, 27,28,29, 32], other prior malignancies [19, 23, 25, 28, 30], prior ipsilateral axillary surgery [26, 27, 29], or distant metastases at diagnosis [19, 21, 23,24,25,26,27,28,29, 31, 32] were largely excluded.

Axillary ultrasonography (AUS) was the most commonly used imaging modality for nodal assessment [19, 21,22,23, 25,26,27,28,29,30,31,32] (Table 2). In most studies, positive needle biopsy of suspicious axillary lymph nodes was mandatory to confirm nodal metastases [20, 21, 24, 27,28,29,30,31,32]. Abnormal AUS [19, 22, 25, 26], magnetic resonance imaging (MRI) [22, 25], or positron emission tomography (PET) scan [23] were used otherwise to define clinical node-positive status. After completion of NST, nodal response was reassessed with physical examination only [20, 27] or with the addition of AUS [