Personalized cancer treatments : realities and prospects

An expensive dream or a vision of the future shared by the cancer community?

New drugs targeting oncogene products activated in several human tumors have been developed over the last 10 years, and some have produced spectacular therapeutic responses in patients resistant to conventional treatments. The variety of genetic and molecular abnormalities found in human cancers has demonstrated their great heterogeneity, even within the same tumor, but has also raised hopes of individualized cancer treatment, with each patient receiving drugs inhibiting the action of the oncogenes involved in his or her own tumor.

The initial results of these therapeutic approaches will be presented and discussed, illustrating some of the successes achieved, but also the current difficulties and limitations. Finally, we will attempt to outline a few avenues for the future.

The most spectacular therapeutic results to date concern :

- Imatinib (Glivec) in chronic myeloid leukemia (CML) containing an activated BCR-ABL fusion protein produced by a t(9-22) chromosomal translocation, and in rare gastrointestinal stromal tumors (GIST) expressing a mutated, activated C-KIT oncogenic receptor on the surface of tumor cells;

- Trastuzumab (Herceptin), whose activity has been demonstrated in breast cancers whose tumor cells overexpress the HER-2 oncoprotein;

- Crizotinib, remarkably effective in rare (3-5%) lung adeno-carcinomas presenting a chromosomal translocation activating the ALK oncogene;

- Vemurafenib , which, for the first time, has achieved major tumor regression in metastatic melanomas whose tumor cells carry the V600E mutation of the B-RAF oncoprotein, present in 50% of patients. Unfortunately, these responses, however spectacular, remain fleeting, and virtually all patients relapse after a few months.

These few successes show that :

- they only concern a very limited number of patients (rare tumours or a small number of patients carrying the target anomaly);

- relapses are frequent, and in some cases constant, due to the emergence of tumor sub-clones resistant to the drug used;

- human tumors do not contain a single genetic event involved in oncogenesis (so-called " driver " mutations), but several (from 5 to 15), with a high frequency of rare anomalies for which the production of future targeted drugs remains highly uncertain. In addition, tumor cells contain many other genetic anomalies (from 50 to 150) which, although not necessary for the survival of malignant cells (so-called " passenger " mutations), may play a role in drug resistance as well as in the relationship between the tumor and its environment (immune system, neo-angiogenesis).

The development of new targeted drugs now requires extensive sequencing and molecular screening programs to enrich early Phase I and II trials with patients who actually have the molecular abnormality targeted by the drug.

The administration of a single drug will almost always lead to relatively fleeting results, and the future clearly lies in the study of drugs given in combination, so as to circumvent or bypass resistance.

However, the study of such combinations of agents is currently very difficult, not only because of their cost, but also because most pharmaceutical companies refuse to authorize combination trials before their agent is marketed. This problem is even more complex when it comes to combining drugs produced by competing companies. Finally, therapeutic development is still constrained by existing drug marketing rules, following a series of Phase I, II and III studies, which are effectively obsolete with this type of agent.

However, new prospects are opening up with the development of agents targeting other mechanisms involved in tumor growth, such as DNA repair, certain epigenetic processes, metabolism and immune responses.

Already, very promising results suggest that the combination (or sequential administration) of targeted drugs inhibiting tumour growth, such as inhibitors of certain tyrosine kinases, with other agents intervening at other stages of tumour development, such as the immune response or angiogenesis, will provide much more durable therapeutic responses.

All these considerations would justify the elaboration of a global strategy for the development of molecular medicine, the aim of which would be the genetic characterization of each individual tumor, so as to identify its potential molecular targets for new therapeutic agents. The field of new technologies -including in particular high-throughput sequencing of tumor DNA, using new generation techniques (NGS)- and the development of new bioinformatics and biostatistical tools will rapidly become necessary and easily accessible to enable the selection of the best drugs or drug combinations to give to each patient (probably through cellular modeling tools, and the development of therapeutic decision algorithms).

In addition, the conduct of therapeutic trials of drug combinations and the detailed analysis of resistance mechanisms likely to occur in patients should lead to a profound change in the types of partnerships between the academic and industrial worlds, as well as an updating of procedures for bringing new anti-cancer drugs to market.

Finally, new medico-economic models need to be developed to make this new type of medicine accessible, and to offset the very high costs of new drugs through a better system for allocating these therapies to patients with a real chance of benefiting from them. These new medico-economic approaches should take into account the cost of the technologies (which is set to fall rapidly) and, above all, the gains induced by no longer exposing the majority of metastatic cancer patients to several lines of conventional chemotherapy, which have very little chance of being effective, and induce additional costs and major toxicities. However, this ineffective, toxic and costly therapeutic approach is now the reality of daily practice for cancer patients whose metastases have resisted one or two lines of "standard" treatment.