Published in pharmaphorum Deep Dive
Early targeted therapies, such as the tyrosine kinase inhibitor erlotinib from Roche/Genentech/OSI, offered the promise of precision medicine to treat cancer at a molecular level. Oncologists noted individual “Lazarus-like” responses to erlotinib, but overall effectiveness was moderate in unselected populations. It was not until the discovery that specific activating mutations in the epidermal growth factor receptor (EGFr) conveyed sensitivity to erlotinib that oncologists could identify the patients most likely to respond and target them for treatment while limiting exposure to those unlikely to respond.
The advent of next generation sequencing (NGS) has reduced the time and expense required for molecular diagnostics, allowing oncologists to personalise treatment at the individual patient level. NGS includes a variety of techniques – the details of which are beyond the scope of this article – that allow for parallel sequencing of large numbers of DNA fragments as opposed to the individual gene tests using earlier PCR techniques and FISH. This is leading to increased collaborations between pharmaceutical and diagnostic companies, such as the recent partnerships between Bayer and Pfizer and Foundation Medicine – which has itself been acquired by Roche.
Although great strides have been made in identifying the treatments best suited for oncology patients to maximise their chance of response and minimise the risk of toxicities using multi-gene panels, a variety of challenging questions exist. Addressing these issues will be key to seeing the full potential of large panel molecular testing in oncology care.
Can I access the test and drug?
From an initial multi-billion dollar investment in sequencing the human genome, large gene panels and whole genome sequencing (WGS) tests are now available for less than $1,000. However, NGS is still largely limited to panels of a few dozen to a few hundred genes for most clinical purposes because of their lower cost and faster turnaround time. WGS remains more of a research tool.
Debate continues as to whether or not personalised medicine based on molecular diagnostics improves outcomes in a cost-effective way. Encouraging publications from Intermountain Healthcare show improved progression-free and overall survival with similar or lower costs in patients treated with targeted oncology therapies selected by a molecular tumour board specifically designed to recommend therapies based on the patient’s genetic profile, but more data are needed to reach a consensus. A recent discussion at the International Society for Pharmacoeconomics and Outcomes Research’s 2019 meeting noted the need for real world cost-effectiveness data and not just clinical trial results to convince payers of the value of NGS diagnostics and targeted medications. Companies seeking to develop new products coupled with NGS would be wise to ensure acceptable pharmacoeconomic results can be demonstrated for maximum product uptake.
While coverage is available for some level of molecular testing, it is normally for specific gene mutation tests tied to a specific therapy. Coverage for large panel tests and WGS is less common. Meanwhile, those in emerging markets must often self-pay, limiting the utility of testing and associated medications to a small fraction of the population in these markets. Diagnostic and pharmaceutical companies wishing to access emerging markets for targeted molecular-based therapies must ensure tests can be provided at little or no costs. Turnaround time is also important. If samples in emerging markets must be shipped overseas for analysis, we have seen how the additional time and expense can make these oncologists reluctant to conduct molecular testing.
What do I do with the results?
When results are available from large panels, oncologists often ask, ‘What do I do with the data?’ Some mutations clearly suggest the use of specific medications in specific tumor types. But the lack of products to address more than a small fraction of the mutations, and the large number of variants of unknown significance (VUS), can leave oncologists and patients with a sense of helplessness. Even worse, if mutations suggest a poor patient prognosis but cannot identify a therapy to try, the result is ‘worse than useless.’ Wishing to avoid these situations can cause oncologists to avoid large panel testing to focus on only single genes with actionable mutations.
Even when a test is completed, access to medications can be denied. Indications are tumor-specific. Oncologists who encounter a supposedly actionable mutation in a tumor where the desired product is not indicated face an uphill battle, which they are likely to lose, if they try to get approval for the medication. In the short-term, issues of actionable mutations in non-indicated tumors can be addressed through compassionate use of products by the pharmaceutical company. The development of more molecularly-targeted therapies will increase the number of actionable mutations, but it is not possible to conduct traditional trials for rare mutations in every tumor type.
One way to address actionable mutations in non-indicated tumors is with basket trials. Patients are defined by the presence of specific mutations as opposed to the tumor type. Products targeting the mutation are then tested regardless of the primary tumor type. Results of basket trials have so far been mixed. The French SHIVA trial concluded that using targeted agents for tumor types where the products were not indicated did not improve progression-free survival in heavily pre-treated patients overall, but did notice subsets that may benefit. Early results are now becoming available from the US MATCH trial. Although it is too early to judge the success of this strategy, basket trials hold the possibility of using large scale panels to recruit enough patients to test the effectiveness of targeted medications in tumors where the incidence of the mutation is too low to justify a traditional design.
Initial biopsy or surgical samples allow physicians to determine what mutations exist. But as tumors progress, their genetic pattern changes. Physicians are faced with obtaining new biopsy tissue, a procedure which some patients are hesitant to undergo, and not all tumor recurrences are easy to biopsy. Liquid biopsy – the detection tumour cells, portions of cells, or DNA circulating in the blood – holds the promise of additional insights into the selection of therapies over time. But liquid biopsy is still new and reimbursement is not assured. Concerns are voiced by some about their sensitivity, specificity, and lack of a standard platform. Nonetheless, these types of tests hold the potential to unleash the promise of targeted therapy throughout the course of treatment.
What about patients?
Genetic changes in the tumor can be somatic or germline. Mutations in genes known to cause a genetic predisposition to cancer, such as BRCA1 and APC, result in germline investigations. Single gene tests or small panels will pick up mutations relevant to the cancer of interest, but large panels can pick up others that may have only a modest likelihood of causing cancer. The key need for the practicing oncologist is to give their patient the best care, which can be done without large panels.
Detection of a mutation that increases the risk of a tumor type different from the one the patient has can launch a cascade of events involving genetic counseling, prophylactic interventions, decisions about whether or not to inform relatives about mutations that may increase their risk of developing a malignancy to a modest degree, and more. Finally, with the frequent reporting of data breaches, who would not be concerned that their genetic profiles could be exposed at some point? All of which can make the process of dealing with the patient’s actual malignancy that much more difficult. These are potential difficulties that can limit oncologist’s willingness to use large-scale molecular testing and make patients reluctant to accept such testing even if it is recommended.
Large panel molecular testing and WGS are powerful tools to help develop personalised medicine. In order to reach the promise of better patient care, companies can consider a variety of strategies. The most obvious answer is for diagnostics and pharmaceutical companies to collaborate to better understand which mutations are truly important and develop more targeted medications to make these mutations clinically actionable. Acceptable cost and pharmacoeconomic data – including from patients treated outside of a clinical trial – will be important to ensure market access. New strategies and techniques such as supporting basket trials and liquid biopsy could further illustrate the benefits of molecular testing and targeted therapies. Finally, companies can assist physicians with the best ways to communicate the benefits and potential pitfalls of molecular testing, including facilitating access to genetic counselors for those practices that do not have them.