Review papers with raw data transparency

Visual summary — biomarker map from Jiang et al. (2025)

Below are visual, data-driven breakdowns of the paper's core taxonomy of resistance mechanisms & biomarkers, followed by a focused critique (strengths, gaps, recommended next steps and falsification tests).

Critical appraisal — strengths

• Comprehensive, clinically oriented taxonomy linking mechanisms to measurable biomarkers and detection methods (NGS/ddPCR/ctDNA, IHC, mIF, spatial omics) — useful for trial design and dynamic monitoring Jiang et al. mapping assays
• Accurate emphasis on heterogeneous, polyclonal resistance (secondary KRAS SII-pocket mutations and convergent bypass events) consistent with clinical sequencing reports Zhao et al. Nature 2021 (resistance landscape)
• Practical translation: links specific biomarker-detected events to rational combinations (KRASi+EGFR/MEK/SHP2) and cites early clinical evidence (e.g., adagrasib+cetuximab in CRC) Yaeger et al., adagrasib+cetuximab

Critical appraisal — limitations & blindspots

1. Narrative-review constraints: no primary data or systematic meta-analytic weighting — important because prevalence and predictive value vary by tumor type (NSCLC vs CRC vs PDAC) and context; many cited biomarkers (e.g., KEAP1) require prospective validation KEAP1 meta-analysis
2. Overgeneralization across tumor types: the review sometimes implies biomarkers are pan-cancer (e.g., KRAS amplifications) though evidence shows tumor-context dependence (KRAS G12C outcomes in NSCLC differ from CRC and PDAC) — quantification of prevalence and PPV/NPV is missing; see PDAC KRAS-dominance discussion JCI KRAS in PDAC review
3. Evidence grading missing: the paper aggregates many candidate biomarkers (FGL1, TF, KRT6A, RPS3, CXCL1/5, ALDH1A1) mostly from single studies or preclinical screens; a formal evidence-level table (prospective validation status, cohort sizes, effect sizes) would aid clinical adoption TF / Oncogene
4. Assay practicalities downplayed: sensitivity limits of ctDNA/ddPCR and spatial sampling bias of tissue biopsies need quantitative discussion (limit of detection, VAF thresholds, false negatives) and cost/turnaround implications for real-world monitoring Bettegowda et al. ctDNA sensitivity

Where the paper correctly synthesizes and where it should be more cautious

Concrete recommendations to strengthen the paper's translational utility

1. Add an evidence-grade table for each biomarker (Level 1 prospective validation / Level 2 retrospective multicohort / Level 3 preclinical only), with cohort sizes and effect sizes where available.
2. Quantify assay characteristics (LOD, analytical sensitivity/specificity, required input sample volumes) for ctDNA, ddPCR and NGS panels — this is essential for implementing dynamic monitoring.
3. Provide tumor-type stratified guidance: which biomarkers are validated in NSCLC vs CRC vs PDAC and where extrapolation is unsupported.
4. Outline prospective trial designs (adaptive umbrella trials) that use early ctDNA changes or single-cell/spatial markers as stratification or interim triggers for combination assignment.

Falsification tests — what would disprove core claims

• If prospective trials show that dynamic ctDNA-detected KRAS secondary mutations or RTK amplifications do not predict progression or do not change outcomes when used to guide therapy changes, that would falsify the claim that these biomarkers are clinically actionable.
• If large multicenter cohorts find no reproducible association between co-mutations (KEAP1/STK11) and therapeutic resistance across tumor types, the generalized prognostic claims must be revised KEAP1 prognostic assertion

Suggested immediate experiments (concise)

1. Prospective ctDNA-led adaptive trial: randomize progressing KRAS-G12C patients (on sotorasib/adagrasib) with emergent NRAS/EGFR alterations detected by ctDNA to KRASi+targeted (EGFR/SHP2) vs physician-choice to measure PFS; this directly tests utility of ctDNA as a decision biomarker.
2. Multicenter validation of KRAS amplification in CRC: standardized qPCR/FISH across centers with harmonized IHC for CD8 to confirm prognostic and predictive value and to power a MEK+CDK4/6 biomarker-directed arm.

Short methodological checklist for clinicians/scientists using the review

• For baseline: do comprehensive NGS (tumor tissue) including KRAS allele-specific CN, KEAP1/STK11/TP53, and RTKs.
• For monitoring: schedule ctDNA panel (NGS or ddPCR) every 6–8 weeks during KRASi therapy and at radiographic suspicion — track emergent allele VAFs and new RTK/RAS pathway mutations.
• For microenvironment decisions: implement multiplex IHC/mIF for PD-L1, CD8, CD206 and spatial profiling pre- and on-treatment where feasible.

Concise conclusions and confidence

Jiang et al. offers a useful, clinician-focused synthesis of candidate resistance biomarkers and practical detection modalities for KRAS mutation-targeted inhibitors; its greatest value is mapping mechanisms to assays and combination rationales. However, prospective evidence grading, quantitative assay guidance, and tumor-specific stratification are needed to translate many proposed biomarkers into clinical practice. Confidence in the review's descriptive accuracy is high for synthesized mechanisms (supported by primary literature) but moderate for clinical actionability because most biomarkers remain at retrospective/preclinical validation stages Jiang et al. caution note