The challenging part of this is that once a tumor has collapsed, presumably infiltrating neutrophils will "mop up" and look like granulocytic MDSC's i.e. ly6c mid, ly6g hi, cd11b+.
Anthony, isn't it suspicious that the phenotype for "granulocytic MDSC" and neutrophils is essentially identical? Just call a spade a space. The entire MDSC field is based on few ancient papers with poorly done flow cytometry using Gr-1 antibody. There is no genetic evidence that it is a separate unique lineage. "Left shift", appearance of the band forms of the neutrophils (which express higher levels of Ly6C and lower of Ly6G) and even seeding of the myeloid progenitors to the tumor-bearing organ are well-known phenomena. Properly designed flow panels will help you to clearly identify proper cell types.
I think it is not easily possible to define MDSC only with their phenotypic markers. you need to discriminate from granulocyte according to the suppressive functions as well. If iNOS, arginase and ROS are expressed in these two well known subtypes, we could tell them as MDSCs; otherwise they are not MDSCs and actually are granulocyte and monocyte/MQs.