I’m talking about de-escalating surgery. There are several things that have caused us to look at de-escalating surgery; one is the recognition that although multimodality treatment has improved patient outcomes the added benefit of each individual treatment is often quite small but the toxicity or the burden of treatment is cumulative. The other thing we’ve learned is that giving adjuvant systemic therapy has a big impact on the outcome of local treatment; it reduces the risk of local recurrence. That leads us to ask the question can we decrease the intensity of local therapy in women who are getting systemic therapy. Finally, we’ve also learned that bigger surgery does not cure bad biology.
So there are three good examples of de-escalating therapy. The first is in ductal carcinoma in situ where the Society of Surgical Oncology, ESTRO and ASCO, the radiation oncology and medical oncology groups, did a joint consensus conference on margins in DCIS in women undergoing lumpectomy with whole breast irradiation. What we found, based on a meta-analysis and review of the literature, is that margins bigger than 2mm did not improve outcome. So since 50% of American surgeons endorse margins bigger than 2mm and European surgeons tend to favour even larger margins than American surgeons, this has the potential to decrease the amount of surgery for a substantial number of women. When we did a similar consensus conference on invasive cancer we found that within two years of publishing it rates of second surgeries had gone down by 16%.
So that’s an example in DCIS; in invasive breast cancer things are actually even more striking. The ACOSOG Z0011 trial was a landmark study that was first published with a follow-up of 6.3 years and showed that in clinically node negative women undergoing lumpectomy and whole breast irradiation who were found to have metastases in the sentinel nodes that whether you were randomised to a complete axillary dissection or a sentinel node biopsy you had the same rate of local recurrence and the same disease free and overall survival. So people were very concerned about that trial, they felt that the follow-up was too short, that the patients were highly selected, it wouldn’t apply to women in general with breast cancer and especially not to high risk younger women, triple negative cancers, HER2 positive cancers.
So now with time we have the answers to most of those questions. Study follow-up was updated this year to ten years; there’s still no difference in the risk of nodal recurrence whether you have an axillary dissection or a sentinel node biopsy alone. There’s no difference in disease free or overall survival. At Memorial Sloan Kettering we started managing patients according to that study prospectively back in 2008; we now have a cohort of 790 patients that we’ve managed that way and we’ve found that we’re able to avoid axillary dissection in 85%, a really big number. If you look specifically at women under 50, those with triple negative breast cancer or those with HER2 over-expressing breast cancer there’s no difference, you can still safely do that. When we look at treatment by radiation fields, whether you use standard whole breast tangents, whether you treat the patient prone or whether you treat the patient with breast and nodal irradiation, local recurrence in all those groups is exceedingly low and not statistically different. So what that tells us is that the Z0011 results apply to the vast majority of women with early stage breast cancer and that they don’t all need nodal irradiation.
The final area where we can de-escalate surgery is in women who are receiving pre-operative chemotherapy, neoadjuvant therapy. The big controversy there is in women who present with metastases in the axillary nodes and appear to downstage clinically, can you avoid axillary dissection by doing sentinel node biopsy? What we’ve found from three prospective single arm studies is that if you find three or more negative sentinel nodes that that accurately tells you with a less than 10% false negative rate that the rest of the axilla is negative. So, yes, you can use sentinel node biopsy in that circumstance. In our experience we found that you could avoid axillary dissection in 49% of patients by treating them that way.
Then the final interesting question is what’s better to avoid axillary dissection – giving chemotherapy preoperatively or applying the Z0011 results and operating on them first? It turns out that the correct answer to that question depends on whether you’re doing a lumpectomy or a mastectomy and whether the patient is ER/PR positive or HER2 positive or triple negative. So for ER/PR positive patients who are undergoing breast conserving surgery if you were to give them neoadjuvant therapy first you would more than triple their chance of needing an axillary dissection, not a good outcome, but for HER2 positive patients having a mastectomy by giving pre-op chemotherapy you reduce the risk of axillary dissection by about 80% and the same thing is true for triple negative patients having mastectomy.
So management of the axilla has become increasingly complex in the modern world. We need to take into account what breast operation the patient is having, what the ER, PR and HER2 status is and whether they have clinically positive or clinically negative nodes to individualise the treatment approach.