Personalised medicine heralds new era for clinical trials cancer drugs

Crizotinib is an ALK (anaplastic lymphoma kinase) inhibitor still in clinical trial that has shown remarkable efficacy in blocking the ALK activating effect of an abnormal chromosomal translocation activating the ALK gene.

The abnormal translocation occurs in only 5 per cent of NSCLC patients and can only be identified in a specialised laboratory; its presence predisposes the carrier to the disease in the first place. The “agony” the patient described, was that, although her chromosomal abnormality made her eligible for the study, it did not guarantee she would be randomised to receive crizotinib, and she was not.

However, following a full course of conventional chemotherapy and having demonstrated sufficient disease progression, she was allowed to cross over the study to receive, finally, the drug that would have been good to have had at the outset. This case illustrates certain frustrations and inadequacies in the clinical trial design of targeted therapies.

The definition of a targeted therapy is a medicine that blocks the growth of cancer cells by interfering with specific targeted molecules needed for carcinogenesis and tumour growth. Targeted cancer therapies are likely to be more effective than conventional cytotoxic cancer treatments and less harmful to normal tissues.

It has been recognised for several years that the traditional design of a clinical trial, running through the classic phase I to phase III endpoints, is not optimal for the testing of targeted therapies. In the early days, potentially useful agents were discarded due to negative trial findings because the patients who were most likely to benefit were not appropriately selected.

Changes have been made, and the new trial designs have moved away from the regular population based model and are referred to as translational research. Translational research is a way of thinking about and conducting scientific research to make the results of research applicable to the population under study and not the general population.

A phase I study conventionally looks for the maximum tolerated dose (MTD). For targeted therapies, these agents are less toxic and work effectively at much lower doses; therefore the aim is to establish a biologically effective dose as opposed to the highest dose tolerated. Investigators attempt to evaluate changes in molecular markers that occur if the desired pathway is affected by the treatment [2].

A phase II study for a cancer agent would look for tumour shrinkage. However, many of the targeted drugs do not cause shrinkage but may create lasting remissions by arresting further growth. Adaptive designs are being explored, such as “randomised discontinuation”, where the drug is administered to all patients and only discontinued following disease progression. This model might especially suit a trial for a disease that progresses rapidly.

Phase III trials have classically used overall survival as the endpoint. It is still necessary to prove that the new treatment is better than the standard therapy and the targeted therapies are no different in this respect. The crizotinib study described above would have been a phase III trial and the novel drug pitched against the standard therapy.

Three steps

In summary, the paradigm for cancer drug development has shifted for targeted therapies to the following steps:

1 Develop a drug to interfere with the function of a molecule readily identified in cancer cells

2 Establish that this target plays a critical role in the abnormal growth and/or invasiveness of the tumour

3 Expect that by targeting this molecule you are likely to create a favorable therapeutic index for the drug [3]

Alongside these goals, a sufficiently inexpensive and robust diagnostic test needs to be developed.

Translational clinical trial design has given us some significant new drugs for the treatment of cancer. Most notably, trastuzumab for human epidermal growth factor receptor 2 (HER2)-positive breast cancer, erlotinib and gefitinib for epidermal growth factor receptor (EGFR)-positive NSCLC, imatinib for chronic myeloid leukaemia (CML) and rituximab for CD20-expressing haematological malignancies.

Clinical trial design for targeted therapies still needs to move another step. Recent advances in genetic profiling are showing that this step is on the threshold of a reality able to be translated to the clinic. There are already commercially available assays that assist in identifying patients with breast cancer who are unlikely to gain additional benefit from receiving chemotherapy. However, the potential to expand the use of this technology is immense.

More extensive genetic profiling could further assist in the optimisation of clinical trials for targeted therapies. The publication in Nature on 16 April 2012 of a landmark study elucidating 10 distinct breast cancer subtypes [4] has shed light on the reasons why some patients, classified according to the conventional criteria of oestrogen receptor (ER), progestrogen receptor (PR), HER-2 status and TNM staging do significantly worse than expected. This work has exploded the future arena for targeted therapy in breast cancer.

Moreover, if patients are routinely genotyped and willing to have their genotype and disease details stored, it should be possible to form a control arm of a study from a retrospective set of data. This would allow patients with rare genetic mutations, such as an abnormal ALK chromosomal translocation, an opportunity to receive the targeted agent at the outset. Surely this is a result that is personally better for the patient and a better use of patient resources.

Libby Hardy is principal pharmacist, cancer services, Royal Devon and Exeter Hospital and Lead Pharmacist, Peninsula Cancer Network

References

1          Back to work after radical Cancer Treatment. Cassandra Jardine. Daily Telegraph, 8 April 2012.

2          CISN Cancer Information and Support Network; Clinical Trial Phases for Targeted Therapies

3          Maitland ML, Schilsky RL. Clinical trials in the era of personalised oncology. CA: A Cancer Journal for Clinicians 2011;61:365–81.

4          Curtis C, Shah SP, Chin SF et al. The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups. Nature;(18 apr 2012): [epub ahead of print] doi 10.1038/nature10983.