Is Ki-67 expression routinely performed on ER+ breast cancer patients and do you find that information useful for making treatment decisions?

Our institution's pathologists do not routinely test for or report Ki-67 expression in any subtype of breast cancer; while I sometimes see this value on path reports from outside institutions in patients referred for consultation or second opinion, I am hesitant to use it to determine treatment recommendations and prefer assays with more retrospective, and now prospective, prognostic and predictive data in ER+ patients, such as Oncotype and Mammaprint. While I have heard of the combination of ER, PR, HER2 and Ki-67, sometimes referred to as IHC4, being used as a low-cost alternative to Oncotype, and might consider doing so in a situation where Oncotype is not available, I am concerned about problems with reproducibility of available Ki-67 assays, which experts are trying to address. The only situation in which I consider using the Ki-67 value to make treatment decisions is in patients started on neoadjuvant endocrine therapy; if the Ki-67 on a repeat tumor biopsy after around 4 weeks of treatment may be predictive of which patients are going to benefit from continuation of this treatment (if the Ki-67 is <10%) as opposed to switching to chemotherapy (if the Ki-67 is >10%).