There is a genomic revolution upon us that is going to reshape the way we approach the treatment of rare cancers.
One of the historical challenges in developing therapies for rare cancers is the dearth of research investment afforded to their study. It’s simple economics. When the patient population is low (and rare diseases are usually defined by a prevalence of one in 1,500 to 2,500 individuals), there is less incentive for industry to address the unmet need, because there is less opportunity to recoup investment in the research and development of the treatment. While the Orphan Drug Act was introduced in 1982 to attract pharmaceutical investment with enticing incentives—such as seven years of exclusive marketing rights—critical voids in our molecular knowledge about rare cancers remain, and as a result, most patients with these diseases are still left in need of safe and effective therapies.
But the rapid evolution of inexpensive genomic technologies is democratizing cancer research strategies globally, with the entire field now recognizing the opportunity for molecular characterization at a “pan-cancer” level, as opposed to the historical single disease bias. This has already led to a number of important clinical discoveries, including the recognition that traditional histopathological classifications of tumors are diagnostically error-prone, and to the equally important identification of distinct molecular differences and commonalities between cancer indications. Such findings hold significant implications for patient treatment, both at the level of therapy selection and in the time it takes to develop new therapeutic approaches to unmet medical needs. This is particularly true for rare cancers.
Despite the enormous medical need associated with rare cancers, experience tells us that they demonstrate some of the highest response rates once a specifically targeted therapy is developed. Associated drug targets are often the result of more discrete genomic aberrations, such as gene fusions, and the resultant fusion proteins are of direct therapeutic relevance, especially if one of the partners contains an enzymatic or small molecule binding domain to which a drug can bind.
It has been suggested that such mutations are associated with discrete intracellular pathways, thus limiting the potential for drug resistance mechanisms. This may account for the clinically significant patient responses and low relapse rates—probably also due to lower levels of tumor heterogeneity. This contrasts with the situation in common cancers, where many pathways can potentially lead to histologically identical diseases. This could explain their typically lower response rates to targeted therapies and the slow and often minimal improvements in overall survival, despite the multiples of investment in these diseases. So despite the historical reticence displayed by the industry, there is significant rationale for intensifying our study of rare malignancies.
Success Stories
The drug imatinib (marketed by Novartis as Gleevec or Glivec) has not only emerged as the poster child for targeted therapy, but the story of its development demonstrates the value associated with researching and developing drugs for rare diseases. Informed by the discovery of the Philadelphia chromosome mutation and the associated bcr-abl fusion protein, imatinib’s development was based on rational drug design. Using high-throughput screening, the investigators used chemical libraries to identify the lead compound 2-phenylaminopyrimidine, which was then modified by the introduction of methyl and benzamide groups to provide improved specificity of imatinib. The clinical response rates seen in chronic-phase CML patients were so significant that the original investigators—Druker, Lydon, and Sawyers—received the Lasker-DeBakey Clinical Medical Research Award in 2009 for “converting a fatal cancer into a manageable chronic condition.”
But imatinib’s success story doesn’t stop there. When it was discovered that tumors from gastrointestinal stromal tumor (GIST) patients often possess activating mutations in the c-KIT or PDGFRA, imatinib again came into focus. GIST is highly resistant to standard chemotherapy and radiotherapy, so the advent of imatinib, which also binds to KIT and PDGFRA, revolutionized the treatment of metastatic GIST, with a clinical benefit rate greater than 80 percent in the metastatic setting and a median survival of 57 months. This simple example clearly demonstrates that initial successes in a single rare cancer setting can open up a bounty of clinical opportunity especially in the presence of molecular commonalities between diseases.
The more recent example of crizotinib further extends this point. Approved by the FDA in 2011, crizotinib has shown significant clinical benefit in NSCLC patients whose tumors harbor the ALK-EML4 fusion protein. Given that ALK mutations are also thought to be important in driving the malignant phenotype in about 15 percent of cases of neuroblastoma, it is now under active investigation in the young patients who suffer from this particular molecular subtype of the disorder.
Such success stories suggest that the molecular and genomic study of rare cancers is sure to reveal many new driver aberrations, some of which will be tractable by available therapies. So I am confident that we will see some quick wins in terms of approaches for treating some of these cancers. In other cases, the findings will spur on drug development efforts to develop unique therapies or arm companies with the knowledge that allows them to go and look at their current therapeutic assets and ask whether any of them might target these novel mutations.
Because these cancers are rare, and because there are typically very poor if any treatment options available, patients with rare cancers are perfect candidates for molecular testing. With many rare cancers, you quickly reach a situation where there is no clinical consensus regarding the best treatment modality, so molecular information offers an exciting avenue for finding an effective cancer treatment. I think rare cancer cases are examples of where more hypothesis-driven treatments can still be useful to clinicians, even if the evidence is somewhat translational in nature.
Consider a treating physician, for example, who is looking for an appropriate clinical trial for a patient with a rare form of sarcoma who has failed all previously described treatment modalities. Armed with appropriate computational methods and clinico-molecular databases, computational analyses may predict a likely activating mutation in a novel kinase target, for which a number of Phase I/II studies of novel targeted agents are currently in progress. Using a well-founded mechanistic hypothesis such as this to direct a patient to a specific trial is certainly much more rational than a choice based solely on generic inclusion criteria.
Or consider a predicted novel activating mutation in an already well-established cancer drug target, or functionally connected regulatory protein. In this instance a clinician could decide to prescribe an approved drug in an off-label setting, even though the evidence is clearly translational in nature. So just as in the case for imatinib in treating GIST tumors, we may be in a situation where it’s possible to take an existing drug and give it to patients with a rare cancer type and see excellent therapeutic effects.
One of the greatest challenges in moving from a histological to a molecular diagnostic approach to cancer is that it is forcing us to reconsider the term “rare cancer.” The case of ALK-EML4 is a good example, where only about five percent of NSCLC patients actually possess this molecular subtype. Could such molecular observations lead to a proliferation in the number of rare cancer types compared to those defined at a histological level? Much will depend on how we choose to define “molecular subtypes.” For example, while ALK-EML4 positive tumors are clearly molecularly distinct from neuroblastomas possessing an activating ALK mutation, there is clearly molecular commonality at the level of crizotinib’s therapeutic mode of action. Once extended across all other cancer types we may in fact discover that the sum of diverse ALK-activating mutations and associated disease mechanisms are actually much more common than we currently appreciate.
This raises exciting implications for repurposing oncolytic agents across many disease indications, and it is this promise that underpins the continued rapid adoption of molecular testing in a pan-cancer manner.
David Jackson, Ph.D. ([email protected]), is CSO at MolecularHealth and also a blogger for Targeting Cancer. Gleevec and Glivec are registered trademarks of Novartis.