ELCC: NSCLC mutation testing highlights ctDNA’s limitations

Key clinical point: In a real-world setting genetic tests that used ctDNA to find treatment-determining mutations in patients with advanced lung cancer were generally less sensitive than genetic tests using biopsy material from the primary tumor.

Major finding: Mutation assessment results from ctDNA and biopsy DNA showed concordance in 89% of specimen pairs tested.

Data source: Prospective study with paired test results from 1,162 patients treated in seven European countries or in Japan.

Disclosures: The ASSESS study was funded by AstraZeneca. Dr. Reck has been a speaker and consultant to AstraZeneca and seven other companies.

View on the News

Use ctDNA with caution

Mitchel L. Zoler/Frontline Medical News

Dr. Egbert F. Smit

The ASSESS study is the largest evaluation of using ctDNA to test the status of a lung tumor for the presence of a mutation in the gene for the epidermal growth factor receptor. It also gives us insight into the feasibility and effectiveness of using this approach in routine practice, outside of a rigorously designed trial. The result was that mutation testing using plasma specimens to obtain circulating tumor DNA was generally doable but resulted in low sensitivity and a low positive predictive value.

The low sensitivity seen overall in ASSESS likely resulted from the multiple testing reagents, commercially available testing kits, and in-house techniques used by individual laboratories in this international study done in diverse settings. Some of the tests for epidermal growth factor receptor (EGFR) mutations are quite sensitive, and others less so. The specific testing method used matters quite a lot. The heterogeneity of methods used in ASSESS reflects the diversity of what usually happens in real-world practice.

Based on the sensitivity limitations, I currently see using ctDNA to test for EGFR mutations only when this is the only option for mutation assessment because a conventional biopsy of the primary tumor failed to provide enough DNA for EGFR mutation testing. Although it is hard to find specific numbers on how often this situation arises in current practice, it seems to happen roughly 10%-30% of the time. So, for about one-fifth of patients with newly diagnosed or recurrent stage III or IV lung cancer EGFR mutation assessment using ctDNA will be necessary.

The data from ASSESS and other studies show that when ctDNA testing identifies an EGFR mutation you can rely on its accuracy. The problem is when ctDNA testing fails to identify an EGFR mutation. In those situations you need to be cautious as you can’t rely on a negative result as being a true negative. I would also be cautious about using ctDNA testing to follow lung cancer patients, as the clinical importance of finding new EGFR alleles appearing in the patient’s blood over time is not yet clear. We need more studies that follow these patients and changes in their EGFR ctDNA over time to determine the clinical relevance of these changes.

The results from ASSESS add to the existing body of evidence that ctDNA testing to find EGFR mutations is feasible in a real-world setting.

Dr. Egbert F. Smit is a professor of pulmonary medicine at the VU University Medical Center in Amsterdam.He had no relevant disclosures. He made these comments in an interview and as the designated discussant for the ASSESS study at the meeting.


AT ELCC 2015


GENEVA – The tests widely available today that use a cancer patient’s tumor DNA that circulates in the blood to detect a treatment-altering lung cancer mutation vary widely in sensitivity, meaning that using circulating DNA to identify a clinically important mutation should be limited to when biopsying the primary does not yield enough DNA for testing.

“It is important to use robust and sensitive methods” when trying to match treatment to the molecular specificities of each patient’s tumor, Dr. Martin Reck said at the European Lung Cancer Conference.

Dr. Martin Reck Mitchel L. Zoler/Frontline Medical News

Dr. Martin Reck

Results from a study that prospectively compared DNA analysis for mutations in the epidermal growth factor receptor (EGFR) gene in tumor biopsies from 1,162 patients with advanced non–small cell lung cancer with DNA analysis of circulating tumor DNA (ctDNA) from the same patients showed that ctDNA analysis had 46% sensitivity, compared with tumor biopsy DNA, reported Dr. Reck, a thoracic oncologist at the Lung Clinic in Grosshandsdorf, Germany. This high level of false-negative results with ctDNA analysis produced a positive predictive value of 78%. In contrast, ctDNA has a specificity of 97% and a negative predictive value of 90%.

The results from the international study, which included 881 patients from seven European countries plus 281 patients from Japan, highlighted the divergent concordance among EGFR mutation comparisons of ctDNA and biopsy DNA depending on which DNA tests were used. Participating physicians could order whichever genetic analyses they wanted. A subanalysis of the study showed, for example, that among the 138 patients whose tumor and ctDNAs were both evaluated for EGFR mutations using the therascreen test marketed by QIAGEN the concordance rose to 95%, the sensitivity of ctDNA was 73%, and the positive predictive value was 94%, Dr. Reck reported.

The Europe-Japan Diagnostic Study for EGFR Testing (ASSESS) enrolled patients with either newly diagnosed, locally advanced stage IIIA or IIIB metastatic non–small cell lung cancer or patients with recurrent non–small cell lung cancer following surgical resection. Patients averaged about 67 years old, and about 85% had stage IV disease. The genetic analyses identified EGFR mutations in 31% of the Japanese patients and in 12% of those in Europe.

On Twitter @mitchelzoler

Next Article:

Three-dimensional models show superiority of nanoparticles in mesothelioma
   Comments ()

Related Articles