Scientists in the Mullighan Lab at St. Jude Children's Research Hospital in Memphis, Tenn., have been researching the genetic landscape of children with acute lymphoblastic leukemia (ALL) for more than a decade. In recent years, the researchers have extended their investigative reach into the genomics of ALL in adolescents and young adults, or AYAs, typically defined as patients between ages 15 and 39.
“Children with ALL generally have a good outcome. There are small proportions who relapse, but overall their outcomes are usually very good. But with adolescents and adults, as they increase in age, their relapse rate increases and their overall survival rate decreases,” Kathryn Roberts, PhD, Staff Scientist at St. Jude, told Oncology Times.
“We wanted to better understand the genomic landscape in these older patients with ALL and compare that to children to see if there are any differences in what may be contributing at the genomic level to these increased treatment failures in older patients. And then we also wanted to see if we could identify any genetic alterations that may prove to be therapeutic targets in adolescent and adult patients,” continued Roberts. She reviewed genetic differences in ALL between children and adults identified by St. Jude researchers and others during an educational session at the 2018 ASH annual meeting (Hematology Am Soc Hematol Educ Program 2018;2018(1):137-145).
“Obviously, in the older populations you have other confounding factors that are not related to genomics, such as not being able to tolerate the intense chemotherapy that children can tolerate, but that's more of a toxicity issue,” Roberts added. “But we wanted to know if there were any underlying genetic alterations. And indeed, there are differences in the genetic landscape between children and adults, and really across the age spectrum.”
In fact, St. Jude researchers have identified many differences in genetic alterations between children and adults in recent years. While some subtypes had been known from cytogenetic analysis done years ago, “we really refined it in a very granular way using comprehensive sequencing to determine what these differences are,” Roberts said.
For example, in children, the researchers see an increased frequency of certain genetic alterations that are known to be generally associated with good outcomes.
“A good example is the ETV6-RUNX1 rearrangement, which is well-known in childhood ALL. And also patients who have more than the normal number of chromosomes, which is called hyperdiploid ALL. We see these at a frequency of 25 percent, roughly, in children, and each is associated with a really good outcome,” Roberts explained. “Conversely, in adults, we see the prevalence of these subtypes decrease to 5 percent. And then we also see a significant increase in genetic subtypes that are associated with poor outcomes, particularly those driven by kinase alterations such as Ph+ and Ph-like ALL.”
According to data cited by the NCI, ALL is the most common cancer in children, accounting for 25 percent of diagnoses in patients under age 15. Between 1975 and 2010, however, the median event-free survival (EFS) rate for childhood ALL increased from 60 percent to nearly 90 percent in children under age 15, and from 28 percent to more than 75 percent in adolescents ages 15-19. But while nine out of 10 children with ALL who receive intensive chemotherapy are cured, adolescent and adult patients historically have had median EFS rates of only 30-45 percent.
Challenges in treating AYAs with ALL currently include “a lower tolerability for [chemotherapy] treatment in AYAs relative to children, the fact that the quality of clinical trials is better in children than AYAs, and, in general, children have been more prognostic to disease than adults,” said Jack Jacoub, MD, a medical oncologist and Medical Director of MemorialCare Cancer Institute at Orange Coast Medical Center in Fountain Valley, Calif.
The most promising treatment for AYA ALL currently is immune-based CAR T-cell therapy, he noted. Also helping “is an overall better understanding of the heterogeneity of the disease and more tailored therapy, as well as improvements in bone marrow transplant tolerability and outcomes.” To help improve survival rates going forward, he suggested “joint research efforts between pediatric and adult leukemia treatment centers in developing larger, well-designed clinical trials.”
In a related vein, recently published results from the CALGB 10403 study show that a course of treatment designed for children with ALL may also be effective in adults. Led by Wendy Stock, MD, from the University of Chicago Medicine Comprehensive Cancer Center, researchers used a pediatric regimen identical to the Children's Oncology Group study, AALL0232, which called for intensive dosing and prolonged treatments of concentrated drugs and escalating doses of methotrexate and PEG-asparaginase.
As reported by the NCI, from 2007 to 2012, 318 patients with a median age of 24 followed the pediatric regimen in the CALGB 10403 study, which researchers deemed safe for AYA patients based a median EFS of 78.1 months, which was more than double the historical control of 30 months; a 3-year EFS of 59 percent; and an estimated 3-year overall survival rate of 73 percent.
Back at St. Jude, researchers in the Mullighan Lab knew that Philadelphia chromosome, encoding the BCR-ABL1 tyrosine kinase fusion, is present in less than 5 percent of children, although its prevalence is more than 25 percent in adults, Roberts noted.
“What we showed is that the frequency of another subtype that we call Philadelphia-like, or Ph-like, which is a subtype driven by additional tyrosine kinase alterations, also increases in adults,” she explained. “So we identified a new subtype that contributed to the genetic difference between the age groups. Through comprehensive sequencing, we also identified other novel subtypes that were previously unknown and characterized their prevalence across the age spectrum as well.”
The subtypes were mostly characterized by rearrangements in new genes, including ZNF384, MEF2D, and NUTM1, Roberts said. “These are transcription factors, and we still need to explore what happens when these genes are perturbed and how that leads to leukemia with functional validation and experimental modeling of subtypes,” she explained.
St. Jude researchers also have identified new subgroups of genes they had previously known were implicated in the development of ALL, or endpoint transcription factors, mainly for a gene called PAX5.
“We had seen these mutations before, but without new transcription sequencing and subtype classification that we developed,” Roberts said. “In one particular study that we published earlier this year... we were able to cluster all of these cases by gene expression profiling and show that they comprise one main subgroup. And that, really, was another new mechanism of subtype classification.”
The St. Jude researchers also identified point mutations in both the PAX5 gene and another gene called IKZF1, Roberts noted.
“Usually, a gene rearrangement is what we think of as the driving element in leukemia development. And all of these other point mutations were thought of as secondary events. They contribute, but they are not initiating. But we saw through our new transcription sequencing classification that, indeed, cases that had a single-point mutation in PAX5 really contributed to the gene expression profile,” she explained.
The researchers also did experimental modeling, introducing single point mutation into mice. “The mouse showed that a single point mutation in a gene could really drive the development of leukemia,” Roberts said. “And that was surprising because we had previously thought of these as mostly secondary events. But through our experimental modeling, we showed these can be initiating events.”
While it is premature to say exactly what clinical implications will stem from the work of the St. Jude researchers, perhaps the biggest thus far is the identification of Ph-like ALL, Roberts said.
“These patients had similar transcription profiles to Ph-positive ALL, but they didn't have the BCR-ABL1 fusion. So it suggested that there may be other kinds of alterations that are driving this phenotype. And indeed, when we did comprehensive sequencing we identified a number of different kinase fusions that collectively make up this subtype,” she said.
The gene expression signature had actually been known for quite some time, Roberts noted. “In Ph-positive cases, they are all characterized by one kinase fusion. But in Ph-like ALL we have identified over 70 different kinase fusions that can all activate the kinase signaling and contribute to this phenotype,” she explained. “And we continue to identify new fusions. So, I think that the genetic heterogeneity that exists within this Ph-like ALL subtype was definitely very surprising. We thought of maybe 1, 2, or 3, maybe 10, but it has been such a genetically diverse subtype.”
The advent of next-generation sequencing, or RNA sequencing, made it possible to identify the subtypes, Roberts noted, adding that the St. Jude researchers also have done a lot of experimental modeling of different types of second-generation kinase fusion inhibitors with “the potential to revolutionize the treatment of this disease.”
“Even though Ph-like ALL has a high number of different kinase fusions, they all tend to converge on one or two signaling pathways, either JAK-STAT or the ABL signaling pathway,” Roberts said. “Even with all of the different fusions, we have shown that they can be targeted with a limited number of approved kinase inhibitors.”
Clinical trials testing kinase inhibitors in children and adults have been initiated at a number of sites, including St. Jude and also the Children's Oncology Group, which is a broad cooperative group for children in the U.S., Roberts said. “The hope, if you will, is that we will see improvement of response and survival with the addition of tyrosine kinase inhibitors to current chemotherapy regimens in patients with Ph-like ALL similar to what we have seen with Ph-positive ALL.”
“We have a good grasp on the new genetic subtypes that we have identified. We know their outcomes and a lot of them are kind of intermediate outcomes,” Roberts continued, in assessing the clinical implications of the research at St. Jude to date.
“Even if you only have chemotherapy to treat these patients, stratification will still be useful in the clinic to improve and refine risk. They can say, ‘OK, maybe these patients should get intensified chemotherapy upfront,’” she stated. “So, I think the clinical implication is that we can still refine and improve stratification in these patients, even if we don't have new therapeutic strategies right now.”
Additional research will focus on the role of novel fusions identified in ALL in children and adults, and most importantly, identifying new potential therapeutic targets within those subtypes,” Roberts said.
“We do need experimental modeling to find out if there any signaling or other pathways that are altered that present themselves as therapeutic, so identifying these and testing new treatment strategies in these subtypes is needed and important,” she said.
“Also, as far as the genomic landscape goes, we have refined the classification of ALL with our comprehensive studies over the past few years, but there still remains about 10-15 percent of cases that lack a known genetic driver, at least by the methods with which we interrogate these cases,” Roberts added. “We call them ‘the others’ because we don't really know what is driving these cases. And if we don't know what is driving these cases, and we only have chemotherapy to treat them, then a large number of patients relapse.
“I think comprehensive sequencing at the whole genome level, and also at the epigenomic level—so that you know what's going on with chromatin modifications—needs to be performed in order to really resolve the genetic basis of these cases and to identify what is driving some of them,” she concluded.
Chuck Holt is a contributing writer.