Skip to main content

Efficacy of different platforms in detecting EGFR mutations using cerebrospinal fluid cell-free DNA from non-small-cell lung cancer patients with leptomeningeal metastases.

Chiang CL, Ho HL, Yeh YC, Lee CC, Huang HC, Shen CI, Luo YH, Chen YM, Chiu CH, Chou TY

Publication: Thorac Cancer (2023)PMID: 36977550View on PubMed

Purpose

This paper compared the detection of epidermal growth factor receptor (EGFR) mutations in cerebrospinal fluid (CSF) of patients with non-small-cell lung cancer (NSCLC) suspected of having leptomeningeal metastases using droplet-digital PCR (ddPCR) and Cobas real-time PCR.  Mutations found in CSF specimens were compared to those found in case-matched CSF cytology and plasma specimens.

Conclusion

The Cobas real-time PCR assay returned “valid” results more often than the ddPCR assay, but the ddPCR assay detected mutations (including common mutations) in a higher percentage of all patients and patients with abnormal CSF cytology results than the Cobas assay.  While the positive predictive value and specificity for a known mutation were 100% for both assays, the sensitivity, negative predictive value and accuracy were higher for the ddPCR than the Cobas assay or CSF cytology (details not provided).  Of the 32 patients evaluated, only 11 patients had detectable mutations in plasma (Cobas assay) and concordance between EGFR mutations detected in CSF and in plasma was only observed in 9 patients. Patients that received Osimertinib after CSF sampling or with lower EGFR fractional abundance (<65%) had a longer median survival than those who did not, but the differences were not significant.

 

Studies (1)

Purpose

This paper compared the detection of epidermal growth factor receptor mutations in CSF of patients with NSCLC suspected of having leptomeningeal metastases using droplet-digital PCR (ddPCR) and Cobas real-time PCR.  Mutations found in CSF specimens were compared to case-matched CSF cytology and plasma specimens. CSF was collected by lumbar puncture from forty patients with EGFR mutant positive NSCLC with suspected leptomeningeal metastases. CSF was immediately centrifuged at 400 g for 5 min at 4°C and the CSF supernatant was recentrifuged at 2000 g for 10 min at 4°C and stored at -80°C. K2EDTA blood was obtained from 32 of the patients by venipuncture and plasma was separated by dual centrifugation at 2000g for 10 min at 4°C. Circulating tumor DNA (ctDNA) was extracted from CSF using the Cobas DNA Sample Preparation Kit and from plasma using the QIAamp Circulating Nucleic Acid Kit. A total of 42 mutations were detected in ctDNA from CSF and plasma using the mutant allele specific real-time PCR-based Roche EGFR Mutation Test.  ctDNA mutations in plasma and CSF were also detected using the TaqMan EGFR T790M Mutation Assay, EGFR Exon 19 Deletions Assay, EGFR L858R Mutation Assay, and EGFR C797S Mutation Assay on a QX200 Droplet Digital PCR system. The 400 g centrifugation pellet was examined by a cytologist.  

Summary of Findings

Thirty of the 40 patients evaluated were found to have leptomeningeal metastases and an additional seven had probable leptomeningeal metastases. The Cobas real-time PCR assay returned “valid” results more often than the ddPCR assay (95.1% vs. 78%, respectively, P = 0.04). Compared to the Cobas assay, the ddPCR assay detected mutations in a higher percentage of all patients (75.6% versus 81.1%) and patients with abnormal CSF cytology results (81.1% versus 86.5%). Importantly, detection of common mutations (exon 19 deletion and exon 21 L585R mutation) was also higher using ddPCR than the Cobas assay (94.3% vs. 77.1%, respectively, P= 0.047). While the positive predictive value and specificity for a known mutation were 100% for all assays, the sensitivity, negative predictive value and accuracy were higher for the ddPCR (94.3%, 50%, and 94.6%, respectively) than for the Cobas assay (75.6%, 23.1%, and 77.3%, respectively) or CSF cytology (analytical details not provided, 78.1%, 25%, and 79.6%, respectively).  Of the 32 patients evaluated, only 11 had detectable mutations in plasma (six with L858R, 2 with exon 19 deletion, 2 with exon 19 deletion and T790M, and one with L858R and T790M). Concordance between EGFR mutations detected in CSF by ddPCR and in plasma using the Cobas assay was only found in 9 of 32 patients (28.1%).  A mutation associated with Osimertinib resistance was identified in the CSF of one patient that developed leptomeningeal metastases after Osimertinib treatment. Patients that received Osimertinib after CSF sampling or had a lower EGFR fractional abundance (<65%) had a longer median survival than those who did not (9 versus 3.5 months and 12 versus 5.4 months, respectively) but the differences were not significant.

  • Biospecimens

    FluidCerebrospinal Fluid
    FluidBlood
    CellCerebrospinal Fluid
  • Preservation Types

    Frozen
  • Diagnoses

    NeoplasticCarcinoma
  • Platforms

    DNADigital PCR
    DNAReal-time qPCR
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    PreaquisitionOther drugs
    No Osimertinib
    Osimertinib after
    Osimertinib before
    Biospecimen AcquisitionBiospecimen location
    CSF cytology pellet
    CSF supernatant
    Plasma
    Digital PCR SpecificTechnology platform
    Cobas real-time PCR assay
    ddPCR assays