Introduction: KEYNOTE-667 (NCT03407144) evaluated pembrolizumab (pembro) plus chemotherapy consolidation in children and young adults with cHL and slow early response to frontline chemotherapy. We present an exploratory analysis of ctDNA in the high-risk pediatric cohort of KEYNOTE-667.

Methods: Patients (pts) aged 3-17 years with newly diagnosed stage IIEB-IVB cHL received 2 cycles of induction with OEPA followed by early response assessment (PET, MRI/CT). Rapid early responders (RERs; Deauville score [DS] 1-3) received nonstudy therapy. Slow early responders (SERs; DS, 4 or 5) received consolidation therapy with 4 cycles of COPDAC-28 + pembro 2 mg/kg up to 200 mg IV Q3W followed by late response assessment (LRA; PET, MRI/CT). Pts with PET positivity at LRA (DS, 4 or 5) received involved-site radiotherapy (RT; 28.8 Gy); pts with PET negativity received no RT. All SERs received maintenance pembro IV Q3W ≤17 cycles. Blood samples for ctDNA analysis were collected from ~150 pts at screening, during induction (C1D28), and after completion of induction (C2D28) and were collected from all SERs (n ≈40) during consolidation (D28 of C1, C2, and C4) and at discontinuation of pembro and from the first 40 RERs during consolidation (end of C1, C2, and C4) and ~8 months after completion of nonstudy treatment. ctDNA from plasma was analyzed using a custom, hybrid, capture-based, targeted next-generation sequencing assay designed by the Institute of Pathology (University of Giessen) for lymphomas (PHLv3; complete coding regions of 136 genes, mutation hotspot exons of 12 genes, and 9 genomic regions frequently affected by translocation breakpoints in B-cell lymphomas).

Results: BaselinectDNA data were included from 38 RERs and 43 SERs. The most commonly mutated genes at baseline were SOCS1, STAT6, TNFAIP3, NFKBIE, and IGLL5 (≥20% of pts). Median [range] maximum variant allele frequency (mVAF) at baseline was numerically higher in SERs than in RERs (10.6% [0.0-34.0] vs 6.2% [0.0-34.2]; AUROC, 0.61 [95% CI, 0.49-0.74]). Among SERs, all pts achieved objective response, but baseline ctDNA status did not differentiate complete response (CR; n = 31; median mVAF: 10.1% [range, 0.0-34.0]) from partial response (PR; n = 10; median mVAF: 8.2% [range, 0.0-20.9]; AUROC, 0.45 [95% CI, 0.25-0.65]) or LRA PET status of negative (n = 29; median mVAF: 10.6% [range, 0.0-34.0]) from positive (n = 12; median mVAF: 5.6% [range, 0.0-20.9]; AUROC, 0.39 [95% CI, 0.20-0.58]).

ctDNA data during (C1D28) and after (C2D28) induction were available for 34 RERs and 37 SERs. During induction therapy, there was a general rapid decline in ctDNA between baseline and C1D28 and a trend toward further reduction by C2D28; a larger proportion of RERs had complete ctDNA clearance vs SERs. Among the 26 RERs who were ctDNA positive at baseline, 24 of 26 (92%) had ctDNA clearance at C1D28 and 26 of 26 (100%) had ctDNA clearance at C2D28; among the 32 SERs who were ctDNA positive at baseline, 16 of 32 (50%) had ctDNA clearance at C1D28 and 21 of 32 (66%) had ctDNA clearance at C2D28. At the end of induction (C2D28), there was a trend towards higher mVAF in SERs vs RERs (AUROC, 0.66 [95% CI, 0.59-0.74]). There was no clear association among SERs between C2D28 ctDNA levels and best objective response (BOR; CR vs PR: AUROC, 0.45 [95% CI, 0.26-0.65]) or LRA (PET negative vs positive: AUROC, 0.42 [95% CI, 0.25-0.59]).

ctDNA data during consolidation and maintenance were available for 11 RERs and 27 SERs. Most pts were ctDNA negative at all timepoints post-induction. Among RERs, no pts were ctDNA positive at start of consolidation, C1, C2, or C4; 1 pt became ctDNA positive at the 8-month follow-up time point. Among SERs, 11 pts were ctDNA positive at start of consolidation: 2 at C1, 2 at C2, 1 at C4, and 1 at the 8-month follow-up time point.

Conclusions: A trend towards higher ctDNA levels in SERs vs RERs was observed at both baseline and end of induction. Among SERs, no association was observed between baseline or post-induction ctDNA level and BOR or LRA PET status. During induction, there was rapid clearance of ctDNA from baseline, with a larger proportion of RERs achieving complete ctDNA clearance vs SERs. During consolidation and maintenance, all but 1 RER was ctDNA negative across all timepoints. Most SERs who were ctDNA positive before entering consolidation achieved ctDNA clearance during maintenance. Results add to the body of evidence indicating that ctDNA may be a promising biomarker in cHL.

Disclosures

Mauz-Koerholz:Merck: Other: Research funding to my institution. Vinti:Amgen, Neovii, Takeda: Other: Travel, accommodations, expenses; Amgen, Takeda: Speakers Bureau; Merck Sharp & Dohme: Research Funding; Amgen, Clinigen: Consultancy. Valero-Arrese:Serb: Other: Speakers' fees; Bayer: Other: Support for meeting registration. Leblanc:Bristol Myers Squibb: Honoraria. Edmondson:Merck: Current Employment. Zhao:Merck & Co., Inc.: Current Employment, Current holder of stock options in a privately-held company. Li:Merck & Co., Inc.: Current Employment. Pillai:Merck & Co., Inc.: Current Employment, Current holder of stock options in a privately-held company. Cristescu:Merck & Co., Inc.: Current Employment, Current holder of stock options in a privately-held company. Peña:Merck & Co., Inc.: Current Employment, Current holder of stock options in a privately-held company.

This content is only available as a PDF.
2024
Sign in via your Institution