No abstract

Purpose Computed tomography is the standard method by which pulmonary nodules are detected. Greater than 40% of pulmonary biopsies are not lung cancer and therefore not necessary, suggesting that improved diagnostic tools are needed. The LungLB™ blood test was developed to aid the clinical assessment of indeterminate nodules suspicious for lung cancer. LungLB™ identifies circulating genetically abnormal cells (CGACs) that are present early in lung cancer pathogenesis. Methods LungLB™ is a 4-color fluorescence in-situ hybridization assay for detecting CGACs from peripheral blood. A prospective correlational study was performed on 151 participants scheduled for a pulmonary nodule biopsy. Mann-Whitney, Fisher’s Exact and Chi-Square tests were used to assess participant demographics and correlation of LungLB™ with biopsy results, and sensitivity and specificity were also evaluated. Results Participants from Mount Sinai Hospital (n = 83) and MD Anderson (n = 68), scheduled for a pulmonary biopsy were enrolled to have a LungLB™ test. Additional clinical variables including smoking history, previous cancer, lesion size, and nodule appearance were also collected. LungLB™ achieved 77% sensitivity and 72% specificity with an AUC of 0.78 for predicting lung cancer in the associated needle biopsy. Multivariate analysis found that clinical and radiological factors commonly used in malignancy prediction models did not impact the test performance. High test performance was observed across all participant characteristics, including clinical categories where other tests perform poorly (Mayo Clinic Model, AUC = 0.52). Conclusion Early clinical performance of the LungLB™ test supports a role in the discrimination of benign from malignant pulmonary nodules. Extended studies are underway.

1091 Background: Triple negative breast cancer (TNBC) remains the most aggressive subtype of breast cancer in which up to 1/3 of all patients will relapse early and distantly. There remains no proven method to monitor and detect early recurrence. This study examines the utility of serial liquid biopsies using a multi-template approach (identification and analysis of cell-free DNA (cfDNA) and circulating tumor cells (CTCs)) as a solution. Methods: 210 patients (average 53 yrs.) with confirmed diagnosis of TNBC, within 3 years of completion of therapy (mean 25 mos.), and in full remission were enrolled. Liquid biopsies were performed quarterly to look for cancer specific genomic mutations not seen in germ line controls, but in cfDNA and CTCs using a custom 27-gene breast cancer panel. Prior validation of our gene panel demonstrated a false positive rate of 0.001-0.0007% in normal cfDNA and cell-based controls. Results: Each subject had on average 3.0 serial samples collected to date. 169 patients (80.4%) had evidence of mutations in at least one sample. Seventy-five percent of cfDNA and 36% of CTC samples examined have been found to bear mutations. The total number of samples with orthogonally (either between template or between draws) confirmed signal was 16.9%. Seven patients (3.3%) have developed documented evidence of recurrence, one in the axilla and the others with visceral/distant metastases. Of the patients that have recurred, all had a persistent mutation on consecutive blood draws or evidence of genomic mutations in both cfDNA and CTCs. Conclusions: The majority of TNBC patients in our study post-treatment were found to have genomic mutations on at least one liquid biopsy sample – either cfDNA or CTC. But, very few had a persistent presence of mutations in subsequent samples and 96.7% of our patients have remained NED. Unlike patients who remained NED, patients who recurred displayed persistent evidence of mutations on serial draws or in both cfDNA and CTC samples. To date our recurrence rate is lower than predicted. This study highlights the need for serial monitoring using a multi-template approach to better fully understand patient specific tumor biology and kinetics. The trial is ongoing. Clinical trial information: NCT02639832.