Will Paired-Normal Sample Testing Become the Norm for Molecular Tumor Assessment?

As precision oncology is becoming the norm, genomic analysis of tumors is rapidly guiding clinical decisionmaking. As a result of both more targeted therapeutic options and the recognition of tumor heterogeneity, molecular assessment of tumors is also becoming more comprehensive.

But with the identification of increasing numbers of mutations in each tumor, comes the inevitable questioning of whether all of the mutations are in fact pathogenic. Paired-normal testing, comparing tumor mutations to normal tissue from that individual, can help hone in on the alterations that are most likely to be pathogenic. This comparison of tumor and normal samples can also reveal germline variants that hold potential clinical implications for both the patient being tested, as well as his or her family members.

Currently, most clinical tumor testing does not currently involve analysis of a matched germline samples, but experts predict that will change in the next few years. While paired testing may not be necessary when assessment is contained to just a few well-characterized mutations, like BRAF, matched testing becomes more important as the scale of sequencing increases.

Is Paired Testing Clinically Relevant?
When tumors are analyzed without a matched normal comparison sequence, germline variants are detected in the tumor sample, but it may be diicult to distinguish their origin as a germline mutation. Thus, the presence of germline variants can complicate interpretation and potentially misidentify the true mutational driver of a tumor. This misinterpretation could have clinical consequences, as most germline susceptibility variants are not targetable, experts say.

“You can’t have personalized medicine without precision genomics,” explains Victor Velculescu, M.D., Ph.D., co-founder and chief science oicer of Personal Genome Diagnostics, whose PGDx’s CancerSelect Targeted Gene Profiling Panel provides for sequencing both normal and tumor DNA to accurately identify true cancer-specific changes. “If you are getting inaccurate information, it defeats the purpose of the test. With tumor-only testing you are getting false positives—extra mutations from the germline that are not subtracted out from the somatic-only mutations. The germline mutations were present, but not specific to the tumor.”

How Frequently Are Germline Mutations Identified?
Two new studies provide evidence of how frequently germline mutations occur in oncology patients not being assessed for hereditary cancers. In the first study, published online Nov. 18 in the New England Journal of Medicine, 8.5 percent of children and adolescents with cancer were found to have mutations in cancer predisposing genes. The authors say these findings suggest that comprehensive genomic screening may be warranted on all pediatric cancer patients, not just those with a family history of cancer.

“This paper marks an important turning point in our understanding of pediatric cancer risk and will likely change how patients are evaluated,” said the study’s corresponding author James R. Downing, M.D., President and CEO of St. Jude Children’s Research Hospital, in a statement.

The St. Jude–Washington University Pediatric Cancer Genome Project used next-generation sequencing, including whole-genome (n=595) and whole-exome sequencing (n=456), or both (n=69), to analyze the genomes of 1,120 children and teens with cancer. Patients had a variety of cancers associated with poor clinical outcomes (leukemia, 52.5 percent and central nervous system tumors, 21.9 percent).

In total, 565 cancer-associated genes were analyzed, with an emphasis on 60 genes associated with autosomal dominant cancer-predisposition syndromes and 29 genes associated with autosomal recessive cancer-predisposition syndrome. The remaining 476 genes were chosen based on published evidence of their role in somatic mutations in cancer. Sequence coverage exceeded 10× for more than 95 percent of the coding exons and 20× for more than 85 percent of the coding exons in the genes of interest.

The researchers found 633 nonsilent germline variants in the 60 genes associated with autosomal dominant cancer-predisposition syndromes. Twelve percent were pathogenic, 3 percent probably pathogenic, and 36 percent of uncertain significance.

The 95 pathogenic or probably pathogenic variants were detected in 21 of the 60 genes in 94 patients. P53 was most commonly involved (in 50 patients), followed by APC (n = 6) and BRCA2 (n = 6). Eight children had germline mutations in the adult-onset cancer– predisposition genes BRCA1, BRCA2, and PALB2, which are currently not included in pediatric cancer genetic screening. The highest frequency (16.7 percent) of germline mutations was seen in children with non-central nervous system solid tumors.

Interestingly, family history did not predict the presence of an underlying cancer predisposition syndrome in most pediatric patients. Of 75 pediatric patients with mutations that were deemed to be pathogenic or probably pathogenic, a review of medical records showed that only 12 patients had previously undergone clinical genetic testing. Forty percent of the medical records with a family history indicated a history of cancer.

This study dispels prior belief that the presence of such germline mutations in pediatric cancer patients was extremely rare and tied to children with strong familial cancer history.

As a result of these findings, St. Jude is initiating a new clinical research study, Genomes for Kids, which incorporates next-generation sequencing into the medical workup of every eligible pediatric cancer patient who enters the hospital for treatment. Any identified germline mutations in a cancer predisposition gene will be referred to the new St. Jude Hereditary Cancer Predisposition Clinic.

In a second recently published paper, researchers found that germline variants are also common in adult patients undergoing tumor-normal sequencing. Nearly 16 percent of patients carried presumed pathogenic germline variants (PPGV) in a gene linked to an inherited human disease, according to the study published Nov. 10 in JAMA Oncology.

Targeted tumor sequencing with matched normal DNA was undertaken using a panel of 341 genes (the MSK-IMPACT panel) in 1,566 individuals with advanced cancer. The majority of the PPGVs were identified in genes associated with cancer susceptibility (most commonly BRCA2 [n=31], CHEK2 [n = 23], MUTYH [n = 23], and BRCA1 [n = 21]). However, the hereditary cancer findings were associated with the individual’s cancer type in only 81 of 198 cases (40.9 percent).

“These findings indicate that it will not be uncommon to detect unexpected actionable variants. Even restricting reporting to the ACMG-endorsed gene set would identify potentially actionable mutations in at least 5 percent of our patients,” write the authors led by Kasmintan Schrader, Ph.D., from Memorial Sloan Kettering Cancer Center in New York. “[Additionally,] over 60 percent of patients carried six or more variants of uncertain significance in genes linked to genetic disease. … Based on the experience reported here, reporting laboratories will require substantial resources for manual variant curation to ensure the quality of their reports.”

Prepping Laboratories
“Clinicians and laboratories should not shy away from the use of genomic technologies in the care of their patients because of the potential to uncover unanticipated information,” write the members of the Clinical Sequencing Exploratory Research Consortium in an article published online Nov. 21 in the Journal of the National Cancer Institute. “Instead, oncology providers, including both ordering clinicians and testing laboratories, should acknowledge the fact that tumor-only testing may reveal actionable germline information and actively implement solutions that maximize the clinical utility of this germline information while minimizing patient misunderstanding and harm.”

The consortium’s Tumor Working Group oers recommendations for how laboratories can prepare themselves for the discovery of germline findings when performing tumor analysis on large panels or exome-/genome-scale sequencing. They say that even in tumor- only sequencing, tumor mutation patterns (hypermutated tumors or massive chromosome rearrangements) may indicate a germline origin to the variant.

“Given the collaboration necessary in variant analysis and clinical interpretation, we believe that both ordering clinicians and laboratories share the responsibility for identifying and managing the potential for germline findings and clinicians will need to appropriately prepare patients for this possibility,” writes the consortium’s lead author Victoria Raymond, from University of Michigan, Ann Arbor.

Among the consortiums recommendations are:

• Carefully construct sequence analysis and data filtering algorithms with an ability towards dierentiating germline and somatic variants.

• Decide what downstream testing will be performed and how it will be performed, when potential germline findings are identified. “Does the lab that carried out the tumor analysis also perform this germline test, and if so how should a germline sample from the patient be collected and the germline test ordered?” Raymond prompts in the article.

• Consider when to recommend referral to a genetic counselor or medical geneticist.

• Recognize the potential for medical, legal, and ethical complications in instances of misclassification and/or misreporting of somatic versus germline variants.

• Develop a standardized approach to the reporting of potential germline findings. The consortium suggests including a precise description of the variant (the specific nucleotide change and the genome build utilized), information for the ordering clinician about the potential implications of a germline finding, and a plan for confirmation, as appropriate.

“Those medical centers oering testing must establish appropriate pretest education tools to inform patients about the potential for identifying inherited susceptibility or previously undiagnosed genetic disease and consider mechanisms for eliciting patient preferences regarding the communication of such results,” writes Schrader from the Memorial Sloan Kettering study. “Medical centers oering such genomic analysis will also need to develop appropriate post-test result communication protocols, as patients currently undergoing tumor mutation profiling may be diicult to engage in traditional post-test genetic counseling owing to their advanced disease. The clinical cancer genetics community will be challenged to establish best practices for communicating results to family members, particularly if a patient with advanced disease who is undergoing tumor-normal sequencing dies before receiving his or her test results.”

Raymond tells DTET that the majority of cancer centers are not currently running paired normal testing because of logistical, turnaround time, and cost considerations, as well as not wanting to deal with additional consent and variant reporting issues.

Takeaway: Analysis of the cancer genome is most informative when paired with a second, normal DNA sample. This practice of distinguishing between variations of somatic and germline origin is expected to be incorporated into comprehensive molecular analysis in the next several years and can impact both treatment decisions for the presenting cancer phenotype and for longer-term surveillance for possibly unrelated cancer syndromes in both the index patient and their relatives.