Scholarlink

Purpose <p>To test, under a pre-planned clinical framework, whether simple multiparametric MRI biomarkers show candidate associations with selected molecular alterations in primary central nervous system lymphoma (PCNSL), with a focus on non-invasive enrichment for CD79B status and the MCD-like phenotype of potential relevance to Bruton tyrosine kinase (BTK) inhibitor stratification.</p> Methods <p>The public UCSF-PCNSL dataset (146 patients; 63 with UCSF500 cancer gene panel) was analyzed. Four clinically motivated MRI biomarker-to-genotype pairings (CD79B, MCD-like, TP53, and CDKN2A/B) were defined in advance of any statistical testing. Effect sizes were reported as Cliff’s delta with 95% bootstrap confidence intervals, and one-sided p-values were adjusted for multiple testing across the four hypotheses at a false-positive threshold of 0.05. Reproducibility was examined by refitting each test after removing one scanner batch at a time (six batches total), and by benchmarking the habitat biomarkers against simple volumetric predictors using cross-validated AUROC.</p> Results <p>The principal hypothesis (H1) showed a higher enhancing-core (h1) fraction in CD79B-mutant patients (Cliff’s delta + 0.34, 95% CI + 0.04 to + 0.63; rank AUC 0.67; <i>p</i> = 0.015; direction preserved in 6/6 scanner batches). H3 was directionally consistent: higher FLAIR/T1c ratio in TP53-mutant patients (delta + 0.31; <i>p</i> = 0.049). MCD-like and CDKN2A/B hypotheses were not supported. After correction for multiple testing, no test reached the 0.05 threshold.</p> Conclusion <p>In this exploratory analysis, enhancing-core fraction on routine MRI emerged as the leading candidate signal for CD79B status in PCNSL, with FLAIR/T1c ratio as a directionally consistent secondary signal for TP53. After correction for multiple testing across the four pre-specified comparisons no test reached the 0.05 threshold; the findings should therefore be considered hypothesis-generating. External multicenter replication, ideally with full LymphGen classification, is required before these biomarkers can be considered for clinical use.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Candidate MRI biomarkers for CD79B status in primary CNS lymphoma: an exploratory radiogenomic analysis of the UCSF-PCNSL cohort

  • 期刊论文

  • 发表日期：

    2026年5月29日

查看全文
  • Clemente García-Hidalgo,
  • Audena María García Kempf,
  • David Galdo Galián,
  • José Antonio Consentino Hernández,
  • José Vicente Cayuela Espí

摘要

Purpose

To test, under a pre-planned clinical framework, whether simple multiparametric MRI biomarkers show candidate associations with selected molecular alterations in primary central nervous system lymphoma (PCNSL), with a focus on non-invasive enrichment for CD79B status and the MCD-like phenotype of potential relevance to Bruton tyrosine kinase (BTK) inhibitor stratification.

Methods

The public UCSF-PCNSL dataset (146 patients; 63 with UCSF500 cancer gene panel) was analyzed. Four clinically motivated MRI biomarker-to-genotype pairings (CD79B, MCD-like, TP53, and CDKN2A/B) were defined in advance of any statistical testing. Effect sizes were reported as Cliff’s delta with 95% bootstrap confidence intervals, and one-sided p-values were adjusted for multiple testing across the four hypotheses at a false-positive threshold of 0.05. Reproducibility was examined by refitting each test after removing one scanner batch at a time (six batches total), and by benchmarking the habitat biomarkers against simple volumetric predictors using cross-validated AUROC.

Results

The principal hypothesis (H1) showed a higher enhancing-core (h1) fraction in CD79B-mutant patients (Cliff’s delta + 0.34, 95% CI + 0.04 to + 0.63; rank AUC 0.67; p = 0.015; direction preserved in 6/6 scanner batches). H3 was directionally consistent: higher FLAIR/T1c ratio in TP53-mutant patients (delta + 0.31; p = 0.049). MCD-like and CDKN2A/B hypotheses were not supported. After correction for multiple testing, no test reached the 0.05 threshold.

Conclusion

In this exploratory analysis, enhancing-core fraction on routine MRI emerged as the leading candidate signal for CD79B status in PCNSL, with FLAIR/T1c ratio as a directionally consistent secondary signal for TP53. After correction for multiple testing across the four pre-specified comparisons no test reached the 0.05 threshold; the findings should therefore be considered hypothesis-generating. External multicenter replication, ideally with full LymphGen classification, is required before these biomarkers can be considered for clinical use.