An investigation into the potential use of serum Hsp70 as a novel tumour biomarker for Hsp90 inhibitors.

Hsp90 inhibitors are under investigation in multiple human clinical trials for the treatment of cancers, including myeloma, breast cancer, prostate, lung, melanoma, gastrointestinal stromal tumour and acute myeloid leukaemia. The pharmacodynamic activity of Hsp90 inhibitors in the clinic is currently assessed by Hsp70 induction in peripheral blood mononuclear cells using Western blot analysis, a method that is laborious, semiquantitative and difficult to implement in the clinic. Since Hsp70 was reported to be secreted by tumour cells and elevated in sera of cancer patients, serum Hsp70 has been evaluated as a potentially more robust, easily and reproducibly measured biomarker of Hsp90 inhibition as an alternative to cytosolic Hsp70. A highly sensitive and specific electrochemiluminescent ELISA was developed to measure serum Hsp70 and employed to evaluate Hsp70 levels in both ex vivo and xenograft samples. In ex vivo studies, maximal secretion of Hsp70 by tumour cells was observed between 48 and 72 h after exposure to Hsp90 inhibitors. In in vivo studies a 3-4-fold increase in serum Hsp70 was observed following treatment with BIIB021 in tumour-bearing mice. Strikingly, secreted Hsp70 was detectable in mice transplanted with human tumours but not in naive mice indicating a direct origination from the transplanted tumours. Analysis of clinical samples revealed low baseline levels (2 - 15 ng ml(-1)) of Hsp70 in the serum of cancer patients and normal donors. Together these findings in laboratory studies and archived cancer patient sera suggest that serum Hsp70 could be a novel biomarker to assess reliably the pharmacological effects of Hsp90 inhibitors in clinical trials, especially under conditions where collection of tumour biopsies is not feasible.