Circulating tumor DNA (ctDNA) dynamics associate with treatment response and radiological progression-free survival (rPFS): Analyses from a randomized phase II trial in metastatic castration-resistant prostate cancer (mCRPC).

Abstract

5508 Background: ctDNA can inform on prognosis, treatment response and survival. We evaluated ctDNA in serial plasma samples from patients enrolled in A.MARTIN (NCT01485861), a randomized phase II study of abiraterone with or without ipatasertib in patients with mCRPC. Methods: Blood was collected in cell-free DNA Streck tubes from 216 patients at 3 time points; baseline, C3D1 and end of treatment. Cell-free DNA (cfDNA) was extracted from plasma using a Circulating DNA Kit (Qiagen) on a QIASymphony machine (Qiagen). 25ng of extracted cfDNA was used in library preparation, constructed with a custom designed, 58 gene, QIAseq Targeted DNA panel (Qiagen) enriched for PI3K/AR pathway genes. Samples were sequenced to mean depth of 3394x on a NextSeq500 machine. Unless otherwise noted, all analyses combine patients across the 3 study arms, and reported p-values are unadjusted. Results: Baseline (BL) ctDNA positivity correlated with radiological progression-free survival (rPFS; HR: 1.8 [95% CI 1.3-2.6], p < 0.01); this association with rPFS was maintained in a multivariate cox model with > 5 baseline clinical variables (HR: 1.6 [95% CI 1.1-2.4]; p = 0.011). Patients with a C3D1 reduction in ctDNA had superior rPFS compared to patients with a C3D1 increase in ctDNA (HR: 2 [95% CI 1.3-3.2], p < 0.01). The rate of ctDNA clearance at C3D1 was higher in the Ipatasertib 400mg arm compared to placebo (56.3% versus 24.4%, p < 0.01). We find that changes in ctDNA associated with best confirmed overall response (p = 0.024); CR patients had the greatest reduction in ctDNA (mean of -23.4%), followed by PR (-16.3%), then SD (-4.1%), and lastly PD patients (-1.3%). Changes in ctDNA levels correlated with SLD changes (rs = 0.289, p = 0.05), and also PSA changes (rs = 0.33, p < 0.01). Changes in ctDNA were associated with rPFS in a multivariate cox analysis that included PSA change (p < 0.01), as well as in a separate multivariate analysis that included SLD change (p < 0.01). Lastly, we explored CNVs and observed emerging resistance mutations in progression samples, including alterations in TP53, AR, FOXA, PTEN, and PI3K/AKT pathway genes. Conclusions: ctDNA analyses may help (i) identify poorer prognosis disease at baseline, (ii) inform on treatment response (CR/PR/SD/PD) and radiological progression free survival (rPFS) in on-treatment (C3D1) samples, and (iii) can elucidate emerging resistance mechanisms at disease progression. Clinical trial information: NCT01485861 .