DYNAMALK: DYNAMic ct-DNA in ALK+ NSCLC Under the Australasian Thoracic Cancers Longitudinal Cohort Biobank Study (AURORA).

Abstract

Introduction: Dynamic ctDNA-profiling in advanced ALK+NSCLC enables personalised treatment. DYNAMALK assesses (1) ctDNA changes relative to prior therapies, PET and tissue biopsy, and (2) resistance patterns to evaluate ctDNA's real-time impact on treatment decisions beyond standard care. Method(s): DYNAMALK- (ACTRN12623000226606p) is a multi-centre, prospective study within AURORA-(ACTRN12625000038493), conducted by the Thoracic Oncology Group of Australasia (TOGA) with consumer codesign. ALK+ patients start first-line lorlatinib (Arm A), or subsequentline of clinician choice (Arm B). Real-time ctDNA sequencing (Guardant360 USA) is performed baseline, first imaging, and disease pro gression, with Molecular Tumour Board recommendations. FDG-PET scans accompany ctDNA time-points, with optional tissue. Analysis is conducted after Arm B completed accrual. Result(s): DYNAMALK opened August 2023, expected to complete enrolment July 2025. In March 2025, N=24/30 are enrolled in Arm A, N=20/20 in Arm B. Median age was 57; 22/44 (50%) Caucasian; 14/44 (32%) had brain metastases. All in Arm A and 4 (20%) in Arm B had a tissue biopsy at entry. In matched cases available, tissue had complete concordance with ctDNA genomic alterations reported in 6/27 (22%), with addi tional alterations detected in ctDNA in 15/27 (55%). Median turn around time for results was 11 days (6-30), with 6/77 (8%) samples unanalysable. In Arm B, 15/20 (75%) received Lorlatinib, 5/20 (25%) Neladalkib. Median of prior lines was 1 (1-3). Arm A, 20/23 (87%) had detectable ctDNA at baseline: 1/23 (3%) with ALK-fusion only; 16/23 (70%) ALK-fusion with non-ALK co-mutations and 3/23 (13%) had non-ALK co-mutation(s) only. Arm B, 16/20 (80%) had detectable ctDNA at baseline: 2/20 (10%) with ALK-rearrangement; 7/20 (35%) with ALK-rearrangement and co-mutations (ALK and non-ALK) and 7/20 (35%) non-ALK co-mutations only. In Arm A TP53 was most frequent, followed by loss of tumour suppressors (ATM, PTEN and RB1), EGFR amplification in 2/23 (9%). MYC amplification occurred in both arms. In Arm B, ALK kinase domain mutation in 4/20 (20%). Of cases available, Arm A, 8/19 (42%) had complete plasma clearance at first time point 2-3 months and Arm B, 3/13 (23%). All patients had baseline FDG-PET. 15/18 (83%) Arm A, 13/14 (93%) Arm B had on treatment FDG-PET matched with ctDNA. All had metabolic response in Arm A, and all but one in Arm B. In both arms complete metabolic response did not correlate with ctDNA clearance. At interim analysis, median follow-up was 6 months (range 1-15) for Arm A and 9 months (range 2-17) for Arm B. Median time-on-treatment was not reached, with 90% (18/20) and 87% (13/15) remaining on treatment. In Arm B, 2 (10%) PD, one had co-occurring MYC-amplification and emergent MET-amplification. The second PD, ctDNA reported the ALK-rearrangement only. Conclusion(s): DYNAMALK demonstrates the feasi bility of a real-world national ctDNA study. This preliminary analysis confirms ctDNA can detect heterogeneous genomics associated with drug resistance. Despite complete metabolic responses on FDG-PET, ctDNA clearance did not directly correlate. Longer term follow up will provided information on molecular evolution of ALK-rearranged NSCLC, receiving next-generation and novel therapeutic sequencing and may help inform treatment personalisation. Acknowledgefunders- LFA, TOGA, ALK+, RCA, Pfizer, G360.