What is the age cut-off to omit radiation therapy after breast conserving surgery for early stage breast cancer?

I agree with the above answers and would like to expand a bit. What is so special about being over 70 (or over 65 for that matter) when your breast cancer is diagnosed? Not much. Age is a rough surrogate for life expentancy and perhaps also for biologic subtype. But we have SEVERAL trials across broader age groups showing relatively low risk of IBTR in women with T1N0 ER+ breast cancer who receive endocrine therapy alone. I summarize these studies below. We also learned that the recurrences trickle in over a protacted timeframe (per annum rates of recurrence stable past the first decade) so long-follow-up is needed.

This is why I believe that many of us are restricing omission of RT to primarily women over 70 with biologically indolent tumors (ER/PR+, H2N-, often gr 1-2). However, studies are ongoing to see if more acurrate assessment of luminal A subtype (Prosigna, Oncotype, ki-67, etc.) will identify younger women with low risk endocrine responsive tumors for whom omission of RT is quite safe as well. I think the days of irradiating every woman <70 after breast-conserving surgery are numbered, but we aren't quite there yet.

Risk of LRR after endocrine therapy alone:

CALGB 9343 (>70, ER+, T1 and cN0): 5 year rate 4%, 10 year rate 10%

PRIME II (>65, <3cm, ER+, no gr 3+LVI): 5 year 5.6% (adding IBTR + regional recurrences)

Canadian - Fyles (>50, Subset analysis T1N0 ER+): 5 year 5.9%

BASO II (women under 70, T1N0, grade 1): 5 year ~4%, 10 year ~8%

ABCSG 8A Trial: (mean age 66, tumor <3cm, gr 1-2, ER or PR+) 5 year 5.1%

The results of the Hughes CALGB 9343 study and the more recent PRIME II data certainly support a discussion and consideration of omission of radiation in the 65-70 year old or above patient population. However, since the long term results from the Hughes study shows a somewhat high absolute risk of recurrence (10% vs. 2%) with omission of RT at 10 years, these results should be discussed with the patient in detail taking into account life expectancy and comorbidities. It is possible that more long term data from PRIME II may support this increased risk over time as well. Considering that there is still a local control benefit over time for these women with radiation treatment , the overall benefit to risk ratio performace status of the patietn should be kept in mind during the discussion and should ultimately provide some context to guide decision-making.

I think all previous studies have clubbed all ER positive breast cancer together, but going forward the phenotype of tumor would be most important in deciding about RT since data suggesting luminal A  type probably have the least benefit. There are ongoing randomised studies looking at omitting RT based on luminal A phenotype (based on ER, PR, and Ki67 level) or low Oncotype score or other biologic test