NEW YORK – In the context of broader and more frequent use of genomic sequencing, molecular pathologists have become more likely — either intentionally or incidentally — to encounter mutational evidence of clonal hematopoiesis, raising new questions about how such signals can, or should, inform clinical decision-making.
On one hand, mutation detection offers hematologists and oncologists what experts argue are useful tools to help them decide how to follow and treat patients with pre-cancer syndromes.
On the other hand, the detection of clonal hematopoiesis (CH) in patients without a hematologic cancer or clear myelodysplastic disorder is also growing — both among solid tumor patients undergoing next-generation sequencing and healthy individuals receiving other types of genetic tests — raising new questions about if and how these incidental findings should be interpreted and/or acted upon.
In a session at last week's virtual annual meeting of the Association for Molecular Pathology, presenters and attendees discussed this rapid evolution and how they view current and future utility.
The term clonal hematopoiesis of indeterminate potential (CHiP) is used to describe CH that is detected in an individual without evidence of hematologic cancer or pre-cancerous myelodysplasia. As genomic sequencing is employed more broadly, CHiP has come under greater scrutiny as an incidental finding and as an important confounder of blood-based, or liquid biopsy sequencing assays.
Research has shown that CH accumulates over time as people age, and that the presence of CHiP portends the development of hematologic cancers, though only in a subset of individuals.
In an effort to find ways to predict which specific CH signals are pathological and which are not, and to hopefully test interventions that might interrupt disease development, several programs have emerged in recent years to follow individuals with CHiP over time and explore strategies for early intervention.
Meanwhile, hematologists and oncologists are increasingly relying on molecular pathology and genomic sequencing to help resolve cases where patients present with cytopenia, a reduction of mature blood cells. In this context, mutation data can help define whether the abnormal blood cell counts are likely to progress to MDS or not, guiding the frequency of monitoring and the potential use of early treatment strategies.
"In the myelodysplastic syndrome world, we are increasingly relying on molecular data to guide us," Amy DeZern, an associate professor of oncology and medicine at Johns Hopkins' Sidney Kimmel Comprehensive Cancer Center, said during the session.
"Molecular profiling … induces anxiety in quite a lot of patients and some providers, both myeloid clinicians as well as hematopathologists … [but I think we need to] really think about using more molecular characterization … because I really think that's the direction that the field needs to go," she added.
Myelodysplastic syndromes (MDS) are what DeZern called a "heterogeneous group of clonal disorders," which represent a precursor state to acute myeloid leukemia (AML) and most often fall under the purview of oncologists despite not being characterized as cancers.
There are relatively few drug treatments for MDS though that has begun to change with a number of ongoing clinical trials including trials of molecularly targeted therapies. However, DeZern cautioned that this "doesn't necessarily mean we should be seeking [more molecular diagnoses] … because we also want to make sure that patients are not receiving therapies inappropriately."
"What we really need to understand at the bedside is how to follow the patient and take good care of them in the context of the biology of their disease," she said.
A workup for a suspected case of MDS is essentially an effort to rule in or rule out the syndrome. Traditionally this is based on the presence of cytopenia, which is a requirement for a diagnosis of MDS. From there, then either the presence of dysplasia on microscopic examination or on certain defined chromosomal abnormalities can cement a diagnosis.
More recently though, mutation data has shown that it can also play a role, at the very least in ruling out, DeZern argued, and in certain cases also ruling in individuals in as either having or being very likely to develop MDS.
"Can the mutational data truly get us purely to a diagnosis of MDS? … The honest answer to that question is not necessarily, but maybe it doesn't matter. We still need the information to understand the biology to help take care of these patients," she argued.
Itself a precursor to AML, MDS has its own precursors: various combinations of cytopenia, dysplasia, and clonal mutation patterns. In the case of an unexplained cytopenia without dysplasia, DeZern said that well-designed NGS panels have a demonstrated high negative predictive value.
Using such tools, a lack of somatic mutations is associated with a risk of MDS as low as 10 percent for suspicious cytopenia cases, even up to 10-year timepoints based on the latest data.
Even in cases where there is some evidence of cellular dysplasia, a lack of mutations might lead a clinician to consider a dysplasia-based diagnosis as a false positive, DeZern added. "A little bit of funny looking-ness under the microscope probably doesn't matter if you don't have any genetic lesions," she said. "There is a caveat that some morphology is always going to be needed, but even still, if you don't find mutations you can feel fairly comfortable [that] even if maybe there's a little bit of a variance under the microscope, you don't need to call it MDS."
In terms of positive predictive value, genomic testing seems to play most prominently in what the field calls "clonal cytopenia of undetermined significance," or CCUS, but which DeZern argued "probably really is just MDS."
"When [patients have] mutations … the probability of evolving to MDS is just short of 100 percent over a 10-year time frame, so in my [mental Venn diagram or] … patient management, I worry far less about any dysplastic features. But if they meet criteria for [CCUS] I follow them very closely," DeZern said.
Moving forward, prospective data will be key as clinicians in this field continue to evaluate and incorporate this emerging data," DeZern said. Some of this is expected to come out of a new effort called the National MDS Study, which aims to establish a cohort of 2,000 adults recently diagnosed with MDS and 500 others with cytopenias that present more of a conundrum.
She also highlighted an abstract slated for the upcoming annual meeting of the American Society of Hematology, which demonstrated that a seven-gene panel — TP53, SF3B1, U2AF1, ASXL1, TET2, STAG2, and SRSF2 — was highly accurate in discriminating MDS in equivocal cases.
As the use of genomic profiling becomes further cemented for patients with known cytopenia, clinical questions regarding those who have CH but no cytopenia or other evidence of disease — in other words, CHiP — are also emerging. These are the cases that arise incidentally from large-scale sequencing initiatives for solid tumor patients, in other comprehensive genomic endeavors, and increasingly through other available genetic testing, including direct-to-consumer options sought out by healthy individuals.
"We are starting to see more people with a clonal hematopoiesis-alone diagnosis," DeZern said.
"They [potentially] come to you when they develop cytopenia, and dysplasia may appear under your microscope or it may not," but when and how these patients progress to a disease that really needs treatment, "that's what we need to understand."
At the same session AMP session, Kelly Bolton, an associate professor of oncology at Washington University, described how CHiP cases that may or may not co-occur with cytopenia are coming to light at greater frequency.
Bolton was previously a leader of the CHiP clinic at Memorial Sloan Kettering, where she said the center's frequent use of comprehensive tumor sequencing in solid tumor patients is now detecting the phenomenon "all the time."
"We developed recommendations for how CH could be handled in solid tumor patients, which is probably where it's detected the most commonly as an incidental finding," she said. Using data on the types of mutations and mutational profiles that seem to confer the highest risk of malignancy, the team developed a strategy whereby low-risk patients are not necessarily recommended to be informed of the finding or worked up further, while those with higher-risk features are considered for additional follow-up.
"If appropriate, then we would recommend that a routine evaluation for occult hematologic malignancy be performed," Bolton said.
However, she added that sometimes a really interesting clinical scenario might present itself where a CHiP finding, absent cytopenia, could still change clinical management for solid tumor patients.
For example, she said "when I was attending at MSK, I had a [patient] in her 70s who presented with her second breast cancer. She had germline testing done that showed a TP53 mutation at 20 percent [variant allele frequency] and we confirmed this was a limited to the blood … through matched testing with fingernails."
"She was considering getting adjuvant therapy for the second breast cancer and we talked a lot about what her risk would be of getting therapy-related [myelodysplasia] given this high VAF TP53 mutation … and to balance the risk-benefit ratio we used [a risk model] to estimate her overall survival benefit from chemotherapy, which turned out to be modest. It was about four percent."
"Since we know that that chemotherapy promotes TP53 mutations, I told her it's very possible that your risk of therapy-related disease could be higher, but we don't really know. We don't have established tests yet," Bolton explained. In consultation with the patient's oncologist, the care team decided on a chemotherapy option with lower myeloid risk.
Looking to the future, Bolton said, it's still unclear how this would work in standard clinical practice, but she and her colleagues are working on better models.
"This isn't ready for clinical action, but I present it as a as an example of how this could be used and how it could actually translate into our clinical practice [as] a really important next step and direction in how we manage CH," she added.
According to Bolton, while these efforts progress, the main consideration right now is patients who turn up with CHiP and also have cytopenia, regardless of which is discovered first. "In terms of clinical trials, it's really only for… that ultra-high-risk group."
For individuals with CHiP alone, "we have nothing to offer them other than recommending cardiovascular screening, which they should be doing anyway," she said.
DeZern agreed, saying that outside of CCUS, CH findings are largely a source of anxiety. Although it's clear the field is going to see changes in this in the near future, "I remain very measured," she added.
Bolton admitted she is already seeing referrals of individuals who discovered they have CHiP via testing from companies like Myriad Genetics or other commercial firms, suggesting that better defining these best practices and cutoffs will become more and more crucial for the hematology/oncology community.
"One of my fears is that with increasing tools in the myeloid arsenal [we risk more overtreatment]," DeZern added, saying that she is aware of experimental protocols now treating patients with chemotherapy or experimental targeted agents for CCUS, or even CH alone.
Bolton agreed, saying that although she is leading this type of research for CCUS, these approaches are thus far only appropriate in the context of clinical trials.
"I think these are slippery slopes. I certainly don't think all patients need chemotherapy, even though I'm an oncologist," DeZern concluded.