Skip to main content
Premium Trial:

Request an Annual Quote

Tumor Heterogeneity Provides Options, Hurdles for Targeted Cancer Treatments


NEW YORK (GenomeWeb) – As investigators continue to develop molecularly targeted cancer treatment methods, strategies for understanding and addressing tumor heterogeneity and clonality are slowly being forced to evolve as well.

Targeted and genome-wide sequencing of tumors has become increasingly common over the past decade, revealing ever more genetic mutations and alterations that might serve as feasible treatment targets. At the same time, genomics has provided a look at the tumor heterogeneity and clonality that can exist in a single individual, either at one point or over time, highlighting cancer's complexity and potential strategies for escaping treatment.

"To some extent, the true malignancy of cancer is reflected in this issue of heterogeneity," said Richard Schilsky, senior vice president and chief medical officer for the American Society of Clinical Oncology (ASCO). "The whole reason, in a sense, that cancer is able to survive and thrive more often than we would like is because it has this enormous capacity to generate heterogeneity, which creates opportunities for it to escape almost any kind of intervention."

That heterogeneity might impact targeted treatment trials in a number of ways, starting with the tumor material used to find targeted mutations, Schilsky explained. The genetic material available for tumors often comes from very small tissue cores or fine-needle aspirate samples, for example. And that means investigators or clinicians are forced to make treatment decisions based on alterations identified in a fraction of one tumor —even when that same tumor looks very different in another section.

"You are making a naïve assumption that what you observe in that tumor biopsy is somehow representative of the total body burden of tumor. And we know that that's not true," he said. "A genomic profile taken from one biopsy might turn out to be different from a genomic profile taken from another tissue biopsy taken one centimeter away from the first one."

Even when mutations of interest are identified, they "don't occur in a vacuum," but in a "molecular milieu" that encompasses other mutations or alterations that may modify a mutation of interest, Schilsky said. "Those types of modifications may not be constant: they're unlikely to be identical in every tumor specimen, if you will, and they're likely to change over time."

Wafik El-Deiry, deputy director for translational research at the Fox Chase Cancer Center, noted that heterogeneity seems to become more widespread in treated tumors, generating opportunities for cancer to escape these therapies and arise in altered forms.

"There's some heterogeneity in treatment-naïve tumors, but not as much as a lot of people might think or expect," he explained. "Once you bring in treatments, there's a significant amount of heterogeneity."

Though such heterogeneity appears to be clinically relevant — potentially contributing to some of the mixed responses observed in individuals treated with targeted therapies — it's difficult to know if, when, and how tumor heterogeneity should be considered over the course of treatment.

If clinicians uncover an actionable mutation in a single biopsy from a patient with cancer, is that enough to warrant targeted treatment? What happens when that mutation appears to be sub-clonal? Is it still worth using that targeted treatment? Will it be clinically beneficial?

"I think we're at a point of trying to figure out, 'Where do we go from here?' in terms of clinical trials that look at heterogeneity and try to show that targeting aspects of heterogeneity will have clinically meaningful effects in how we treat patients," El-Deiry said.

"It's an area that needs some attention," he added. "It needs to be thought about and addressed in clinical trials … and I think some people would ask the question, 'If we were to address this, would it really help the patient?'"

The potential of an alteration in a fraction of the tumor to be druggable might depend on the nature of that mutation, particularly if it is helping to drive cancer formation. Likewise, investigators are establishing a firmer view of the mutations that tend to occur early on in tumor development and are, consequently, shared across clones, though not all of these alterations are currently targetable.

Overall, clinical trials still have a ways to go in dealing with tumor heterogeneity effectively, Schilsky said. Some academic and commercial groups are taking a crack at quantifying the problem by developing algorithms aimed at predicting an individual's response to a drug or combination of drugs based on the genomic profile of their tumor, not just a single alteration, for example, though these need to be validated.

And despite all these considerations, there is evidence that molecularly targeted treatments can outperform conventional treatments, whether or not heterogeneity is accounted for.

"Where there has been a biomarker selection strategy to identify patients who receive the targeted drug, the targeted therapy approach always is superior to chemotherapy," Schilsky said. "Those trials certainly support the notion that, when the target is present and when the drug is known to be an effective drug, that's the preferred approach to conventional chemotherapy."

At the ASCO annual meeting in Chicago earlier this month, he offered an update on the Targeted Agent and Profiling Utilization Registry (TAPUR), a prospective, non-randomized, phase II basket study aimed at assessing the utility of molecularly targeted drugs in tumors with actionable mutations. That study is set to expand to new centers and may soon include additional targeted drugs.

Though TAPUR is not designed to address tumor heterogeneity and clonality effects on molecularly targeted treatment outcomes, Schilsky noted that the study may help in unraveling related issues surrounding, for example, mutation detection assays, biopsy methods, and treatment strategies.

In the meantime, there are at least some clues that targeted treatments are effective even when the alteration they impact is present at very low levels. During a session on tumor and clinical heterogeneity at the ASCO conference, Memorial Sloan Kettering Cancer Center researcher Barry Taylor outlined results from a clinical trial of the ATP-competitive AKT kinase inhibitor AZD5363, for example, that was also described in a study published in the Journal of Clinical Oncology earlier this year.

For that trial, Taylor and his colleagues used AZD5363 to target AKT1 E17K mutations in advanced solid tumors of 52 individuals who had breast, gynecologic, or other types of cancer.

Despite the advanced state of their cancers, and a long list of treatments the individuals had already undergone, the investigators found that the drug stretched out progression-free survival. And, more unexpectedly, individuals with AKT1 mutations seemed to benefit from this targeted treatment even when the mutation was found at a very low frequency or in a minor tumor sub-clone, Taylor noted.

Conversely, there could theoretically be risks associated with ignoring an actionable mutation found at low frequency in a tumor, El-Deiry cautioned, since cells containing these genetic changes may flourish as other, treatment-vulnerable clones wane.

"The whole allele frequency issue is something that's dynamic: it may be less relevant at a certain point, but over time, it may become more relevant for that patient," he said, noting that "a lot of the studies you see on molecularly-targeted therapies are not so much focused on [heterogeneity of] secondary resistance mechanisms of primary targeted therapy."

That's not always the case, though. Joel Neal, a thoracic oncology researcher at Stanford University, pointed to trials that have attempted to thwart targeted treatment escape and rampant tumor heterogeneity before they occur with therapies developed to staunch resistance as a first line treatment.

In a paper published in the New England Journal of Medicine this month, for example, an international team shared findings from the ALEX trial, a phase 3, randomized, open label trial for advanced but untreated non-small cell lung cancer (NSCLC) cases involving ALK gene rearrangements. The trial compared the effectiveness of the ALK-inhibiting drug crizotinib to an ALK tyrosine kinase inhibitor called alectinib that appears to be effective against resistance mutations associated with crizotinib treatment.

At the ASCO meeting, the ALEX trial team presented data for progression-free survival that exceeded 25 months, on average, in the alectinib-treated group compared to just over 10 months in individuals who were treated with crizotinib.

Such results hint that alectinib may prevent some mechanisms of heterogeneity and resistance, noted Neal, who was not involved in the trial. Unfortunately, though, "in a given patient's tumors, there may be more than one of those mechanisms of resistance cropping up at once," he added, making it difficult to generalize resistance-related mutation consequences across patients and cancer types.

"This heterogeneity question really needs to be assessed on a patient-by-patient level," Neal said. "It's the next level of personalized therapy."

Though there's still a ways to go, experts suspect there will be benefit in approaching cancer from an evolutionary point of view, tracking tumor changes with approaches such as sequencing consecutive biopsies or analyzing circulating tumor DNA.

The latter is appealing since liquid biopsies — which can be collected more frequently and less invasively — are also expected to represent a greater proportion of total tumor mutations than any single biopsy, El-Deiry said.

If ctDNA samples are collected intermittently, Neal said, they could provide an opportunity for finding resistance mechanisms and planning a potential treatment strategy before a tumor reappears by standard imaging.

While liquid biopsies "might provide a more holistic view of the mutational load and the heterogeneity of the mutational landscape in an individual," though, he cautioned that "there's almost no way to prove that hypothesis without simultaneously obtaining blood and a tissue biopsy from every identifiable tumor site in a person's body."

Alternatively, some clinicians are attempting to use biopsies to profile genetic patterns in tumors found in a patient after an initial response followed by resistance, though there are limits to the feasibility and desirability of doing biopsies over and over in the same individual.

"There's only so many times you can stick a large needle into someone's vital organs and get away with it," Schilsky said.

Whatever the method used to track tumor dynamics, experts agreed that tumor heterogeneity, clonality, and related resistance mechanisms warrant further attention by those hunting for more widespread and long-lasting cancer therapies.