ESMO 2014 Report: Precision medicine: Panacea or false dawn?

Reports from the European Society for Medical Oncology 2014 Congress in Madrid by Denys Wheatley:  In his talk on Visualising the true spectrum of cancer- will we get convergence from divergence, Peter Campbell (Cambridge, UK) referred to studies on over 1500 AML patients, sequencing 170 genes in toto. This looked for frequency of genes associated with cancer and the possible driver mutations among them.  The spectrum from FLT3 to ATRA, JAK2 and BRAF followed a declining exponential in frequency (from 3 to 10, probably more in patients who had solid tumour). The mapping on an individual patient of mutations gave profiles that covered up to 10 defects, these weighted directional indications giving some indication in prognosis as to the best and worse outcomes for patients. Many variables were examined – age, cytogenetics, white cell counts etc.  – but the conclusion was that a cohort of 300 gave as much information as one of over 1000. The question was whether this helped in terms of more precise diagnosis and treatment, with an answer that did not suggest any particular advance at this stage of analysis. The possibility that transcriptomics would help seemed to more a cry for help to find better correlations.

José Baselga (New York, USA) addressed the question as to Drug registration strategies: how will these change with precision medicine? The four features of the talk related to defining the mutation drivers (referring mainly to lung, breast and ovarian cancers), defining molecular subtypes of tumour, developing appropriate drug platforms, and the registering patients for trials on this basis. With certain mutations, there were no trials set up, e.g. ERBB2 mutations. Diversity makes it difficult to choose which combination of drugs in therapy might work best, with tumour heterogeneity making precision medicine difficult. The dawn does not seem to be approaching if this is the case.

Andrew Hughes (Macclesfield, UK) summed up the session on the basis that cohorts of patients would undergo screening for mutations, from which stratification would become increasingly complex, and this would come down finally to a personalised basis, which might be equated in many ways to what is meant by precision medicine. Following what has to be done to find all the sequencing data, mutations and risk factors that are involved in precision medicine for the individual patient, it is clear that this is not a cheap option. He pointed out the problem with trials in which usually two arms (treated vs controls) would each have to be subdivided into unselected and selected patients. It would be the latter which might provide some information that could advance precision medicine. The size of an arm in trials is highly dependent on the efficiency of the initial screening, as this improves, fewer patients would be needed to get significant results (and the more affordable), but the case for more personalised medicine more clearly arises from them. The conclusion was that we do not yet know if we can reach a panacea or if it is a false dawn. It must be quite evident from the whole session that the word “panacea” is hardly one that can be applicable in oncology.