ctDNA biomarker potential highlighted in CROWN, VISION trials

Author: By Lynda Williams, Senior medwireNews Reporter 

medwireNews: The CROWN and VISION trial updates presented at the 2021 ASCO Annual Meeting demonstrate the potential usefulness of circulating tumour (ct)DNA as a biomarker for treatment response in non-small-cell lung cancer (NSCLC) patients with driver mutations. 

CROWN ctDNA Analysis 

Ross Soo, from the National University Cancer Institute in Singapore, reported changes in ctDNA among advanced NSCLC patients with ALK alterations who participated in the CROWN trial of first-line treatment with the third-generation ALK tyrosine kinase inhibitor (TKI) lorlatinib versus crizotinib. 

The study’s secondary endpoint of biomarker ctDNA was determined using plasma samples taken at baseline and again at weeks 4 and 24 and on completion of treatment. The average variant allele fraction (VAF) for ALK alterations was expressed as a change over the course of treatment, where a delta VAF of below 0 indicates a decrease in ctDNA, the presenter explained. 

In addition, patients were classified as having a molecular response if the ratio of mean treatment VAF to mean baseline VAF was below 50%, with patients considered “cleared” if they had complete clearance of ctDNA by week 4 or 24, he said. 

Ross Soo reported that around half of patients in both treatment arms had ctDNA detected at baseline and these patients had a greater tumour burden than those without. After 4 weeks, tumour burden was reduced in ctDNA-positive patients regardless of treatment arm allocation, although the “difference appeared to be more prominent in the lorlatinib arm”, he commented. 

Most lorlatinib-treated patients with baseline ctDNA achieved “a rapid reduction” in their mean VAF by week 4, with a few more doing so between 4 and 24 weeks, and “very few” had stable ctDNA or an increase in ctDNA after 24 weeks, the presenter continued. 

A similar pattern was reported for crizotinib-treated patients, albeit with more patients experiencing stable or increasing ctDNA by week 24. 

Moreover, there was a positive correlation between the reduction in mean VAF between baseline and week 4 and the reduction in tumour burden, and progression-free survival (PFS) was also significantly associated with a negative change in VAF at week 4, with significantly longer PFS reported for lorlatinib than crizotinib in patients with ALK fusion and/or mutations (median unreached vs 7.4 months, HR=0.62). 

Analysis by molecular response to lorlatinib at 4 weeks showed that compared with non-responders, patients with no ctDNA at baseline and who were “cleared” responders at 4 or 24 weeks had a significantly longer median PFS, with hazard ratios of 0.16 and 0.29, respectively. 

Of note, while patients without ctDNA had the longest PFS in the crizotinib arm, the median PFS was generally shorter than that achieved by any lorlatinib-treated patient and molecular response did not appear to be a useful biomarker for predicting PFS with this agent, Ross Soo observed. 

Concluding that “ctDNA may be an early indicator for lorlatinib efficacy” in this patient population, he said that “ongoing studies are further investigating the relationships observed here and the potential for using dynamic ctDNA monitoring to inform treatment decisions and improve outcomes patients with non-small-cell lung cancer.” 

VISION Exploratory ctDNA Analysis 

Paul Paik, from Memorial Sloan Kettering Cancer Center in New York, USA, reported an exploratory biomarker analysis from the VISION trial of patients with advanced NSCLC with a MET exon 14 skipping alteration who had previously received up to two lines of treatment. 

The trial has previously demonstrated “meaningful clinical activity” of treatment with the selective MET inhibitor tepotinib 500 mg/day among 99 patients who were enrolled on the basis of liquid and/or tissue biopsy compared with the overall study population, he said. This included an overall response rate of 52.5%, a median duration of response of 14 months and a median PFS of 7.4 months, with comparable findings for pretreated and treatment-naïve individuals. 

Plasma samples were taken at baseline, at weeks 6 and 12 of tepotinib therapy and at the end of treatment, and assessed for ctDNA alterations in MET, EGFR, PIK3CA, KRAS and other oncogenic drivers, the investigator explained. 

Paul Paik said that the current baseline analysis confirmed earlier findings indicating that “clinical outcomes were not impacted by the type or location of MET exon 14 skipping alterations” but a “trend towards better efficacy was seen in patients with concomitant MET amplification”. 

Six of the eight patients with MET amplification at baseline achieved an objective response, and one patient without MET amplification but with a concomitant MET kinase domain mutation at baseline also had “a relatively long PFS” of 17.3 months. 

The presenter also pointed out that three responses were achieved among the five patients with a baseline PIK3CA pathway alteration whereas there were no responses among the three patients with KRAS or NRAS mutations. EGFR amplification did not appear to have an effect on outcome but patients with wild-type TP53 had numerically better median duration of response and median PFS than those with a concomitant TP53 mutation. 

Turning to the on-treatment ctDNA findings, Paul Paik reported that 46 patients achieved at least 75% depletion of the MET exon 14 VAF relative to baseline, indicating a confirmed molecular response. These patients had an objective clinical response rate of 76%, a median duration of response of 14 months, and a median PFS of 11 months, “which is longer than what we saw with the overall study population.” 

By contrast, none of the five patients who had an increase in their VAF from baseline achieved an objective response and their median PFS was just 5.5 months. 

Paul Paik emphasized that “the correlation between primary molecular and clinical progression suggests the presence of a clonal resistance mechanism in a subgroup of patients with MET exon 14 skipping NSCLC.” 

Finally, the investigator reported findings from the end-of-treatment ctDNA samples. Among 52 patients who discontinued tepotinib due to disease progression, including patients enrolled on the basis of tissue biopsy MET exon 14 skipping alterations only, 13% had a MET kinase domain mutation not found in earlier ctDNA samples. 

This included seven patients with emerging Y1230 or D1228 mutations who all responded to treatment, with five achieving PFS of longer than 10 months. 

In addition, samples taken from 35 patients at time of disease progression revealed new alterations including MET mutation or amplification (n=7), TP53 mutation or RB1 loss (n=6), EGFR or HER2 amplification (n=4) and RAS/PI3KCA mutations (n=3), but “no high frequency bypass pathway alterations were detected”, the presenter remarked. 

“Liquid biopsy provides a means to monitor response to treatment assess resistance mechanisms, with a view towards refining the therapeutic approach to improve patient outcomes”, he concluded. 

Future Outlook 

Session discussant Lecia Sequist, from Massachusetts General Hospital in Boston, USA, said more research is needed to determine how best to use ctDNA to guide subsequent clinical actions in patients with available targeted therapies. 

“I think we can now begin to see that the utility of ctDNA to match patients to therapies, and to prognosticate outcomes, especially via clearance of ctDNA, is likely a universal principle of driver mutation-positive lung cancer”, she added. 

“We likely do not need to reprove these concepts with each new type of driver mutation cancer, and each new TKI that comes to market”, she continued, suggesting that future effort should be on designing trials to determine how best to use ctDNA to improve outcomes, such as guiding escalation or de-escalation of therapy or use of adjuvant treatments. 

References 

Soo RA, Martini J-F, van der Wekken AJ, et al. Early circulating tumor (ct) DNA dynamics and efficacy of lorlatinib: Analysis from the CROWN study. J Clin Oncol 2021;39(suppl 15; abstr 9011). DOI: 10.1200/JCO.2021.39.15_suppl.9011 

Paik PK, Veillon R, Felip E, et al. METex14 ctDNA dynamics & resistance mechanisms detected in liquid biopsy (LBx) from patients (pts) with METex14 skipping NSCLC treated with tepotinib. J Clin Oncol 2021;39(suppl 15; abstr 9012). DOI: 10.1200/JCO.2021.39.15_suppl.9012