In addition to the various isoform targeted inhibitors developed to date, there is also significant emerging potential for p110???and dual p110??????inhibitors for the treatment of immune inflammatory diseases and cancer and also of p110???inhibitors in the latter therapeutic area. Also important  to demonstrate ADX-47273 target and pathway modulation in both the preclinical discovery phase and the early clinical development of PI3K inhibitors. This is critical in the implementation of the Pharmacological Audit Trail, enabling rational optimization of dose and schedule of administration as well as go/no go decision making. In addition, progress has also been made on the identification of potential predictive biomarkers for the identification of patients that are most likely to respond to PI3K inhibitors. These include PIK3CA mutation, PTEN expression loss, HER2/ERBB2 amplification/overexpression, wild type KRAS and gene expression signatures.
Finally to be highlighted is the emerging picture from the clinic INNO-406 of PI3K inhibitors as generally well tolerated agents that are already beginning to show evidence of single agent therapeutic activity in early clinical trials in cancer patients. Concerns about potential effects on glucose metabolism appear to have been alleviated, with only mild effects being seen, at least with the doses and schedules used to date. What then are the key issues facing the preclinical discovery and clinical development of class I PI3K and class I/class IV inhibitors for cancer treatment? Identifying optimal isoform selectivity profiles has already been discussed and is ongoing. Related to this point, more work needs to be done to identify the best predictive markers of sensitivity for drugs with different selectivity profiles.
In addition, further research on biomarkers of resistance, both intrinsic and acquired, is also essential. At the moment the available biomarkers are probably best described as enrichment biomarkers for use in enriching early clinical trials for patients with malignancies with molecular characteristics that make them more likely to respond. Much more work needs to be done to validate and clinically qualify biomarkers that may be truly predictive. It needs to be remembered that, especially since PI3K inhibitors can have effects on tumour angiogenesis and tumour microenvironmental interactions, there may not be a single biomarker of sensitivity but rather a group of these or a predictive molecular signature.
Studies in preclinical systems, including large molecularly characterised cancer cell panels and human tumour xenografts, together with genetically engineered mouse models, will be useful for this. However, it is likely that many of the answers will be worked out by molecular profiling, including cancer genome sequencing, of clinical tumour material and the correlation of such data with therapeutic response and outcome. Many PI3K inhibitors are now progressing through phase II single agent efficacy studies and the results are eagerly awaited by the oncology community. Combination studies are also underway. Because of the number of potential combinations, it may take some time to identify optimal combinations and both preclinical and clinical studies will be important for this. Some rationally based combinations, for example with MEK inhibitors, have obvious mechanistic appeal and these are being prioritized.