Very interesting trial.
It is suprising to me that the trend toward OS exists, particularly after such short follow-up... especially since no individual trial of BCS +/- RT has ever shown a survival benefit and it took the very large Oxford meta-analysis to demonstrate a 15-yr survival benefit from adjuvant local or locoregional RT.
Moreover, I feel that this data in many ways contradicts the findings of ACOSOG Z0011. That trial suggests that regional nodal treatment (RT OR surgery) is not necessary in pts with 1-2 positive sentinel nodes. Of course, the main limitation of ACOSOG was its failure to accrue... but the 1.6% incidence of locoregional failure is compelling.
Despite the impressive findings from MA.20, I still think that tailored regional nodal irradiation (RNI) is indicated for patients with 1-3 positive nodes. 85% of patients on MA.20 had 1-3 positive nodes, which makes the findings even harder to believe. The findings might shift the subtleties of RNI decsions for me, but won't compel me to treat regional nodes comprehensively for patients with 1 positive node. Below is my "poor man's" synopsis of how I plan to approach these patients in my practice:
MA.20 trial obviously begs the question of who specifically needs RNI and which nodal regions need to be treated. Treatment of the undissected SCV/level III is easy to justify, in my opinion, for patients with 3 positive nodes. These patients weren't eligible for ACOSOG anyway. Median age of MA.20 was young (53yrs in MA.20 vs 60yrs in Z0011) thereby confering higher risk of locoregional recurrence. I routinely treat SCV/level III for 2 positive nodes in young patients already. In Z0011, 71% of patients on the SLN bx alone arm had only 1 positive node, so I may shift my practice to treat SCV/level III for ALL patients with 2 positive nodes based on MA.20 (since Z0011 did not fully accrue and proportion of patients with 2 positive nodes small). Still not eager to treat SCV/level III for 1 positive node in absence of other high risk features (high grade, LVSI, T2-3 primary, etc), as I think that it is probably overkill.
The most difficult questions, in my opinion, are regarding management of the axilla. All SLN positive patients in MA.20 were required to have ALN dissection. Moreover, all patients randomized to RNI had axilla treated with PAB. Questions: Do we need to treat the SLN positive, but undissected axilla? Does this trial say that we need to start adding a PAB or contouring and intentionally treating the low axilla in patients with 1-2 positive SLN in whom ALN Dx is omitted? Or even if ALN Dx is performed? I will probably use high tangents for all with positive SLN bx, who don't get axillary dx (per ACOSOG) in whom I am not treating SCV/level III. High tangents were not explicitly excluded on ACOSOG and was probably done by the majority of the investigators in that trial. I will probably use steeper tangents to cover level I-II axilla (without PAB) with monoisocentric match to SCV/level III for 2-3 positive nodes and undissected axilla. I doubt that I will intentionally boost axilla with a PAB in patients with 1-3 positive nodes if ALN Dx performed, except in setting of gross ECE or if surgeon suspects microscopic residual. I think that the risk of lymphedema with ALN dx plus intentional axillary RT in MA.20 is lower than I have observed in practice (as was acknowledged by David Wazer in an interview about this abstract).
I agree with GFunk that treatment of IMNs for 1-3+ nodes seems like overkill. Therapeutic ratio cannot routinely favor tx of this IMN nodal basin for these patients. Will consider for medial/central primary and >1 positive axillary node, but will probably talk myself out of it the vast majority of the time.
