Andrew D. Zelenetz, MD, PhD
AS MORE is learned about the genomic landscape in non-Hodgkin lymphoma, clinicians are grappling with how to apply this information in the clinic. At the 2018 Pan Pacific Lymphoma Conference, Andrew D. Zelenetz, MD, PhD, helped them understand this emerging area.1 Dr. Zelenetz is Professor of Medicine at Weill-Cornell Medical College, Attending at Memorial Sloan Kettering Cancer Center (MSK), and Chair of the B-Cell Lymphoma Panel of the National Comprehensive Cancer Network® (NCCN®).
Across the lymphoid malignancy landscape, at least 30 different somatic mutations occur at a frequency of at least 5%. Certain ones tend to track with specific histologies. For example, mutated MLL2 is found primarily in large B-cell and follicular lymphomas, whereas TET2 tracks strongly with T-cell lymphoma.
“There are a large number of mutations, but trying to understand the complexity becomes very difficult,” Dr. Zelenetz said. Genomics can provide key information in a number of areas: diagnosis, prognosis, tracking minimal residual disease, understanding clonal evolution, and selection of therapy.
“I think one of the most important applications of genomics in lymphoma is in refining the diagnosis,” he said. “Despite immunohistochemistry, there are still cases that are unclear, since we are dealing with about 80 different entities.”
Which Genomic Panel?
SINGLE-GENE mutation studies have largely been supplanted by a number of different next-generation sequencing panels. These platforms differ from each other in terms of their methods for calling variants, their list of genes, and the speed of turnaround. “The various platforms cannot be used interchangeably,” he emphasized.
For instance, MSK-IMPACT-Heme and Foundation One Heme assays overlap in only about 60% of the genes. Other panels such as the RainDance Technologies’ ThunderBolts Cancer Panel include only 10% of the genes in the larger panels but have a much more rapid turnaround time.
The incorporation of a “matched normal” is a useful feature to eliminate variants of unknown significance, he maintained, and this is found in the MSK-IMPACT-Heme panel but not in the Foundation One Heme panel. The MSK-IMPACT-Heme panel uses a hybrid capture method and targets 400 genes. The Foundation One Heme panel uses a hybrid capture and includes a DNA-coding sequence for 405 genes and an RNA-coding sequence for 265 genes.
“The most important application of genomics in lymphoma is in refining the diagnosis. We are dealing with about 80 different entities.”— Andrew D. Zelenetz, MD, PhD
Tweet this quote
Of particular utility in patients with leukemia are the ThunderBolt Cancer Panel and the Archer RNA sequencing test. The ThunderBolt panel is limited to 53 genes and also lacks a matched normal, but its turnaround time is 4 to 7 days, as opposed to 3 to 4 weeks for other panels. The Archer test rapidly identifies 146 gene fusions, also in a fairly timely manner. These tests can be helpful when a prompt diagnosis and treatment plan are important, which is the case with leukemia, added Dr. Zelenetz. Since different platforms have different strengths and weaknesses, the selection of the genomic test should be driven by the clinical information the user is seeking, he added.
Genomic Testing Informs Diagnosis
DR. ZELENETZ offered a few cases to illustrate how genomic testing can refine diagnosis. One case involved a 38-year-old woman diagnosed with grade 3 follicular lymphoma with transformation to diffuse large B-cell lymphoma (DLBCL). The Foundation One Heme test identified nine genomic alterations in SOCS1, CD58, PCLO, TNFIP3, and CIITA. The fact that CIITA fusions are found in 38% of individuals with primary mediastinal large B-cell lymphoma helped to clarify the diagnosis; the patient did not, in fact, have primary mediastinal large B-cell lymphoma. The presence of the CIITA mutation also was informative for prognosis: the 10-year disease-specific survival is 64% for patients with primary mediastinal large B-cell lymphoma with this mutation, vs 85% for those without this mutation.2
As for this patient, a single mutation can matter, but it sometimes is more complicated than this. In mantle cell lymphoma, for example, all TP53 alterations are not the same: mutations, but not deletions, strongly impact outcomes, explained Dr. Zelenetz.
This finding was established by a retrospective analysis of the Nordic MCL2 and MCL3 trials, which evaluated 183 patients treated with current standard-of-care regimens.3 Patients with TP53 wild-type disease had much better progression-free and overall survival and a lower cumulative incidence of relapse than did those with TP53-aberrant disease (ie, with TP53 mutations or 17p deletions).
“However, the devil is in the details,” Dr. Zelenetz noted. “We actually see different results in different diseases when we look at deletion 17p vs TP53 mutation.” Patients with deletion 17p, in whom one copy of TP53 is missing, had no detriment in overall survival and only a minor impact on progression-free survival, but mutation in TP53 did impact outcomes.
As the authors reported, TP53-mutated cases had a “dismal” outcome, with a median overall survival of 1.8 years, compared with 12.7 years for TP53-unmutated cases (P < .0001). They concluded that TP53 mutations identify a phenotypically distinct and highly aggressive form of mantle cell lymphoma that shows poor or no response to standard treatment. The authors suggested that patients be stratified according to TP53 status and that patients with TP53 mutations be considered for experimental front-line trials.
“Understanding these subtle prognostic differences can be important, because you don’t want to overtreat or not recommend transplant because of deletion 17p,” Dr. Zelenetz said. “If the patient is TP53 wild-type, the outcome is actually going to be pretty good. We have to be careful about the details.”
Levels of Evidence Informative
THROUGH A collaborative effort of the Association for Molecular Pathology, American College of Medical Genetics and Genomics, ASCO, and the College of American Pathologists, guidelines have recently been established, including a ranking of evidence that supports the importance of genomic alterations in diagnosis, prognosis, and treatment.4
The categories are intended to help providers interpret research in clinically meaningful ways.
According to the MSK-IMPACT-Heme platform, the frequency of clinically relevant genomic alterations in DLBCL exceeds 80%, but fewer than 25% of them are supported by Level A evidence. “As the field develops, and we start to get more robust data sets, these levels of evidence will be very helpful to us as clinicians, to know which mutations we should be focusing on,” he said. “You may get 400 gene mutations back. They’re not all created equal.”
Difficulty in Conducting Biomarker-Driven Trials
RECENT EXPERIENCE at MSK has demonstrated the difficulty in conducting biomarker-driven trials based on mutations. Many patients must be screened, even if only a few go onto clinical trials.
"We need to integrate genomic sequencing into routine practice, and this is going to be aided by tools embedded in electronic medical record systems.”— Andrew D. Zelenetz, MD, PhD
Tweet this quote
For example, in a study selecting patients by the presence of CREBBP and EP300 mutations, only 9% of 78 eligible patients actually went on to receive a matched agent on a clinical trial. “This was very disappointing. Why would there be big gaps between identifying the mutation and getting the patient on trial? Well, the CREBBP/EP300 trial was for patients with relapsed DLBCL and follicular lymphoma, but DLBCL patients progress fast, and you don’t have 6 or 8 weeks for a 400-gene panel to come back,” he explained. “These trials are not easy to do.”
Accessing and Interpreting Genetic Data
IN LARGE cancer centers such as MSK, sequencing results may be more accessible at this time than they are for many other practitioners. “We have a unique opportunity in that our genomic information is directly accessible from the electronic health record,” he said. A click of the cBioPortal icon brings up the record of all the available sequencing results. A small “flame” beside a mutation labels it a “hot spot,” indicating its importance in patient care.
Information is also shown from OncoKB, a vetted precision oncology database that collates information from multiple sources about the mutation, potential treatments, and clinical trials. This information is curated by a clinical genomics annotation committee, applying the levels of evidence as described here. “You don’t have to go back to the original source material to figure it out,” he said.
“We need to integrate genomic sequencing into routine practice, and this is going to be aided by tools embedded in electronic medical record systems,” he said. ■
DISCLOSURE: Dr. Zelenetz has received research funding from MEI Pharmaceuticals, BeiGene, Genentech/Roche, and Gilead; has received honoraria for consultation or participation in advisory boards from Amgen, AstraZeneca, BeiGene, Celgene, Gilead, Genentech/Roche, Janssen Scientific, Novartis, Pfizer, Pharmacyclics, Takeda Pharmaceuticals, and Verastem; and is on the scientific advisory board of Adaptive Biotechnologies, Inc.
1. Zelenetz AD: Clinical applications of genomic studies in NHL. 2018 Pan Pacific Lymphoma Conference. Invited Lecture. Presented July 17, 2018.
2. Steidl C, Shah SP, Woolcock BW, et al: MHC class II transactivator CIITA is a recurrent gene fusion partner in lymphoid cancers. Nature 471:377-381, 2011.
3. Eskelund CW, Dahl C, Hansen JW, et al: TP53 mutations identify younger mantle cell lymphoma patients who do not benefit from intensive chemoimmunotherapy. Blood 130:1903-1910, 2017.
4. Li MM, Datto M, Duncavage EJ, et al: Standards and guidelines for the interpretation and reporting of sequence variants in cancer: A joint consensus recommendation of the Association for Molecular Pathology, American Society of Clinical Oncology, and College of American Pathologists. J Mol Diagn 19:4-23, 2017.