When (if ever) would you use fulvestrant up-front instead of an aromatase inhibitor for ER+ or PR+ metastatic breast cancer?

There are 2 scenarios based on the data and clinical guidelines that one can consider fulvestrant as 1st line therapy for ER+ metastatic breast cancer. Denovo or treatment naive stage 4 patients with bone only mets did better on fulvestrant vs an AI in the FALCON trial. SWOG 0226 showed a survival benefit for AI+fulvestrant 1st line in denovo treatment naive stage 4 and is noted as an option in both ASCO and NCCN guidelines. The ASCO practice guidelines recommend considering fulvestrant +/- palbociclib for patients who have recurred on or less than 12 months after adjuvant AI therapy. This would be especially the case if an ESR1 activating mutation is detected in the recurrent tumor or cell free tumor DNA, as a retrospective analysis of SOF...

Fulvestrant is probably as good or better than AI therapy in the first line setting but has less evidence. It's also less convenient with requirement of monthly injections. The situation that comes up most frequently in my practice is if a patient develops metastatic disease while on adjuvant AI or very soon after (less than 12 months). In those situations, I prefer to switch to Fulvestrant +/- a CDk4/6 inhibitor instead of continuing or switching AIs as the backbone of therapy.

I often use combination anastrozole and fulvestrant in endocrine naive patients given a relative 26% overall survival advantage compared to anastrozole alone, or in patients who relapsed >10 years after initial diagnosis (irrespective of prior endocrine treatment), given about a relative 40% progression-free survival advantage, as shown in S0226. The data from FIRST and FALCON also suggests that use of fulvestrant upfront is associated with better outcomes (PFS/OS in FIRST, and PFS in FALCON, OS pending in FALCON), specifically in endocrine naive populations.

In case of S0226, the overall survival advantage was seen for upfront fulvestrant despite cross over to fulvestrant in the anastrozole alone arm (patients were encouraged to cross over ...

The key question is whether there is an advantage for using an AI with fulvestrant upfront, and although S0226 noted that there was, the follow-up PK study from S0226 showed a significantly lower anastrazole levels in the combination arm throughout the treatment course (Hertz, Br J Clinical pharm 2016). This suggests that the predominant effect in S0226 was primarily a fulvestrant effect, rather than a combination effect with an AI. I therefore do not think there is a strong rationale for combining an AI with fulvestrant in the frontline setting, and there is certainly enough evidence to discourage the practice beyond first-line.

One can certainly use fulvestrant as a frontline single agent based on FIRST and FALCON, but I would caution that...