At the AACR meeting on Molecular Diagnostics and Cancer Therapeutics meeting in Denver, there was quite a bit of interesting scientific data coming out on cancer biology and biomarkers, so here is a quick synopsis of what appealed to me:
1. IGF-1R is over expressed in a subset of triple negative breast cancers (TNBC)
This presentation, from Witkiewicz et al., was interesting because they showed that IGF-1R might be a useful prognostic biomarker in TNBC. Usually, TNBC, which occurs in 15-20% of all breast cancers and is associated with a poorer prognosis, so finding a subgroup that might actually do better could be useful. Gene amplification was seen in 23% of the cases investigated. Caution must be exercised here, however, just because something is over expressed or amplified, does not mean that it is mutated, and therefore a potential druggable target with a therapeutic as we saw with the negative phase III results with figitumumab, an IGF-1R inhibitor in lung cancer.
What I would like to know though, is how many women with TNBC also have the BRCA1 or 2 mutation and have amplified IGF-1R? We know that targeting a cancer with one drug at one or two mutations in solid tumours has modest effects. But what if we target several things with a combination therapy, for a specific subset, then that might possibly yield different results altogether.
2. Molecular biomarker analysis using circulating tumour cells (CTCs)
Siminder Atwal, a Genentech scientist, presented her work on CTCs, with the idea of determining whether they could be used as a predictive biomarker by correlating CTCs with HER2 status in archival tumour samples. CTC's in theory should match tumour biopsies since they are cells that have shed from the primary tumour. The Genentech scientists found that they were indeed correlated, with a high concordance (95%) using the CellSearch system. Of course, EpCAM expression can vary in some tumour types, complicating analysis and interpretation, but it looks like they saw some early hints that molecular biomarker status in CTCs are indeed reflecive of the biomarker status in patient tumours.
Predictive biomarkers allow us test whether a patient is likely to benefit from a given treatment, so the obvious and leading question is whether this work could be extended to look at eg CTC's in colorectal, lung or breast cancers to determine whether a patient is more likely to respond to bevacizumab (Avastin) or not. It would probably be easier in breast (lots of biopsy samples) and colorectal (lots of surgery providing samples).
At the moment, there is no way of telling who is most likely to respond to Avastin, so a predictive biomarker test would be really useful for clinicians before deciding on treatment, rather than having to expose thousands of patients to the systemic side effects. Given that the FDA have to make a decision by the year end on the full approval of Avastin in breast cancer, it is a shame that they likely won't have this sort of data to help guide a decision.
3. BRCA1 mutations in prostate cancer
Gerhardt Attard gave an enlightening talk entitled, "Circulating tumor cells: Potential uses, pitfalls and challenges in their use as pharmacodynamic markers" but what struck me was not so much the fascinating information about CTCs as his mention that a subset of prostate cancer patients have the BRCA1 mutation. They tested the impact of olaparib, a small molecule PARP inhibitor from AstraZeneca, and found that all three responded.
A study open for men with prostate cancer and either BRCA1 or 2 mutations is currently recruiting patients. Given the interest with PARP inhibitors in breast and ovarian cancers, I'll be following this development with great interest.