© 2018 Mac millan Publishers Li mited, part of Springer Nature. All rights reserved. therapeutic benefit of treatment with dasatinib and quercetin is comparable to clearing senescent cells in INK-ATTAC mice9. However, it will still be important to determine if the same cell types are targeted in the INK-ATTAC and the dasatinib and quercetin-treated mice. In fact, the mechanism of cell clearance in INKATTAC mice differs fundamentally from mice treated with senolytic agents. In the transgenic models, cells with high p16ink4a promoter activity are targeted, while senolytics act by killing cells expressing specific pro-survival SCAPs. Importantly, not every cell with high p16ink4a expression is senescent and not every senescent cell has high p16ink4a expression. For example, p16INK4A and SA-β-gal activity are induced in macrophages as part of a reversible response to physiological stimuli even though these cells are not senescent10. An additional approach to provide evidence of senolytic activity in vivo involves the use of human tissue explants harbouring senescent cells7. A putative senolytic should attenuate senescence markers while inducing apoptosis. However, while this approach documents senolytic activity of a compound, it does not identify the types of cell undergoing cell death and whether it is only a subset of senescent cells. Additional methods, including flow cytometry and CyTOF-based approaches, are needed to define the cell types within a tissue explant that are targeted by senotherapeutics, as well as to colocalize markers of senescence and apoptosis. Proving that a compound exhibits in vivo senolytic activity is further complicated by potential indirect effects on the immune system. Functional immune cells, including natural killer and T lymphocytes, can remove senescent cells. However, as the immune system ages, the ability to clear senescent cells wanes, contributing to the accumulation of senescent cells. It is possible that some drugs, for example, rapamycin, reduce the burden of senescent cells indirectly by improving immune cell function. Similarly, inhibiting the SASP might improve immune clearance of senescent cells. Thus, demonstrating that a drug directly targets senescent cells must address the possibility that the drug improves host clearance of senescent cells. All in all, there is reason to be tremendously excited about the health and economic impact of senotherapeutic drugs. However, as noted, caution is warranted in interpreting the mechanism of action and relative specificity, selectivity and efficacy. Furthermore, there is still much to learn about how best to identify, characterize and apply these drugs to improve human health.