KEY POINTS
- This Canadian population-based retrospective study examined 89 women with cT1–2N0 breast cancer and at least three positive sentinel lymph nodes after upfront breast surgery between 2010 and 2017. Thirty-two patients received regional nodal irradiation (RNI) without completion axillary lymph node dissection (cALND), while 57 underwent cALND with or without RNI. Patients receiving neoadjuvant chemotherapy or presenting with clinically positive axillary nodes were excluded. Median follow-up was 7.6 years.
- The comparator groups received heterogeneous combinations of surgery and radiation. Of the 57 patients undergoing cALND, 42 also received RNI and 15 received dissection without RNI. Overall, 73/89 patients (82.0%) received RNI, commonly using 16-, 20-, or 25–28-fraction schedules. The most frequently used prescription was 45 Gy in 20 fractions. Among irradiated patients, 89.0% received axillary levels I and II coverage, 98.6% received level III/supraclavicular irradiation, and 95.9% received internal mammary nodal irradiation.
- Axillary recurrence was uncommon but numerically higher with RNI alone. Five-year axillary recurrence was 6.3% with RNI alone versus 1.8% with cALND ± RNI (p=0.26), corresponding to two versus one observed axillary recurrence. Five-year recurrence-free survival was 81.3% versus 75.4% (p=0.53), respectively. Across the entire cohort, 25 patients (28.1%) developed recurrence, including three axillary recurrences (3.4%) and 22 distant recurrences (24.7%). The small number of regional events prevents a reliable conclusion that omitting axillary dissection is oncologically noninferior.
- Overall and breast cancer-specific survival were similar between groups. Five-year overall survival was 87.5% after RNI alone versus 82.5% after cALND ± RNI (p=0.53). Five-year breast cancer-specific survival was 87.5% versus 84.2% (p=0.67). These findings were observed despite most patients having a substantial nodal burden: 66/89 (74.2%) had exactly three positive sentinel nodes, and patients undergoing cALND had a median of one additional positive node identified at dissection.
- Lymphedema was significantly less frequent when axillary dissection was omitted. It occurred in 1/32 patients (3.1%) receiving RNI alone versus 12/57 (21.1%) undergoing cALND ± RNI (p=0.02), an absolute difference of 18 percentage points. Within the dissection group, lymphedema rates were similar after cALND alone and cALND plus RNI (20.0% versus 21.4%; p=0.91). This association is consistent with the recognized morbidity of extensive axillary surgery, although lymphedema assessment included both objective findings and clinician-confirmed symptoms rather than one uniform prospective measurement protocol.
- Treatment selection and baseline differences limit causal interpretation. Mastectomy was more frequent in the cALND group (71.0% versus 46.9%; p=0.02), while patients also differed in treatment era and regional RT fractionation. The dissection comparator combined patients with and without RNI, making it difficult to isolate the incremental oncological benefit of surgery. Furthermore, the 6.3% versus 1.8% regional recurrence difference could be clinically meaningful despite not reaching statistical significance.
- The findings cannot be directly extrapolated from or used to extend existing axillary de-escalation trials. ACOSOG Z0011 excluded patients with more than two positive sentinel nodes, while patients with at least three positive nodes represented a small minority in AMAROS. This study contributes real-world data for that underrepresented population but includes only 89 patients, predominantly hormone receptor-positive tumors, and historical systemic treatment. Randomized evidence with contemporary systemic therapy and standardized RNI is needed before routine omission of cALND in patients with ≥3 positive sentinel nodes.
CLINICAL TAKEAWAY
Regional nodal irradiation without completion axillary dissection was associated with substantially less lymphedema and no statistically significant survival disadvantage in this small cohort of patients with at least three positive sentinel nodes. However, axillary recurrence was numerically higher without surgery, and the study was insufficiently powered to exclude a clinically relevant difference. Omission of cALND in this population remains an individualized decision rather than a validated extension of existing de-escalation criteria.