Analysis & ranking
PHASE 2 — Evidence and Impact Analysis
Article 1 — Ibrutinib in Early-Stage CLL: CLL12 Genetic Substudy (PMID 42322115)
| Dimension | Score | Rationale |
|---|---|---|
| Scientific Novelty | 7 | First large-scale genetic stratification analysis from a prospective RCT in early-stage CLL; the del(17p)/TP53 non-benefit signal is clinically important and not previously established from trial-level evidence in this pre-treatment population |
| Clinical Relevance | 8 | Directly informs treatment decisions: confirms watch-and-wait for all subgroups on OS grounds; identifies ibrutinib-resistant genetic profile (del17p/TP53) in the setting where treatment might have been considered |
| Population Reach | 6 | CLL is the most common adult leukemia in the West (~21,000 new US diagnoses/year); early-stage asymptomatic CLL is a large and growing population due to incidental detection; not a rare disease |
| Implementation Speed | 8 | Genetic testing for IGHV, del17p, TP53 is already standard practice in CLL; findings plug directly into existing clinical workflows; no new infrastructure needed |
| Evidence Strength | 8 | Retrospective genetic substudy of a prospective RCT (CLL12), n=515, 69-month follow-up, published in Blood by GCLLSG — one of the most credible CLL research consortia globally; main limitation is substudy design (not pre-specified per subgroup) |
Key quantitative result: No OS benefit in any subgroup; EFS benefit in U-IGHV, del(11q), +12, NOTCH1, ATM, NFKBIE but not del(17p)/TP53.
External validation: CLL12 is a registered Phase III RCT (EudraCT 2013-003211-22); this is a post-hoc substudy, so subgroup findings are hypothesis-confirming rather than independently pre-registered.
Main limitation: Retrospective genetic substudy — subgroup analyses were not the primary pre-specified endpoint; multiple comparisons inflate Type I error risk across mutation subgroups.
Equity implications: Patients at academic centers with access to comprehensive genetic panels benefit most; community oncology settings with limited FISH/sequencing access may lag in applying these findings. International applicability is high given GCLLSG's European breadth.
Evidence Maturity: ✅ Confirmed Validated — trial-derived, large n, long follow-up, high-impact journal.
Article 2 — Tumour Budding Prognostic Meta-Analysis in Stage II-III CRC (PMID 42322147)
| Dimension | Score | Rationale |
|---|---|---|
| Scientific Novelty | 6 | Tumour budding is an established concept (TNM 2017, WHO 2019); this meta-analysis is the largest to date and adds statistical robustness, but the directional finding is not surprising — incremental rather than paradigm-shifting |
| Clinical Relevance | 8 | Stage II CRC adjuvant chemotherapy decisions are genuinely difficult; a robust independent prognostic marker with HR ~2.5 for DFS could tilt adjuvant treatment decisions for high-budding tumors; directly actionable for GI oncology and pathology |
| Population Reach | 8 | CRC is the 3rd most common cancer globally (~1.9 million new cases/year); stage II-III represent the majority of resectable disease worldwide — very large reach |
| Implementation Speed | 7 | Tumour budding assessment requires no new technology (H&E staining); ITBCC scoring is standardized; main barrier is pathologist training and workflow standardization, which is achievable in 1–3 years at organized cancer centers |
| Evidence Strength | 8 | Systematic review and meta-analysis of 43 studies, n=17,831; large effect sizes are consistent across studies; main limitation is heterogeneity in budding scoring methods across included studies and lack of prospective therapeutic intervention data |
Key quantitative result: DFS HR 2.53 (95% CI 2.14–3.00); OS HR 2.40 (95% CI 1.69–3.42); CSS HR 3.37 (95% CI 2.19–5.19).
External validation: Meta-analysis pools 43 independent study populations; internal consistency of effect across stage II and III is reassuring, though not a prospective therapeutic trial.
Main limitation: Heterogeneity in tumour budding assessment methodology across contributing studies; no included RCT directly tests whether budding-guided adjuvant therapy improves outcomes.
Equity implications: H&E-based assessment is low-cost and globally accessible in principle; however, ITBCC standardization and pathologist training resources are unevenly distributed — high-income settings will implement faster. Low/middle-income countries with high CRC burden may lag without active training programs.
Evidence Maturity: ✅ Confirmed Validated — largest meta-analysis to date, consistent large HRs, methodologically sound.
Article 3 — MBD1 Aberrant Splicing Drives MDS Epigenome (PMID 42322117)
| Dimension | Score | Rationale |
|---|---|---|
| Scientific Novelty | 9 | Genuinely novel mechanism: MBD1-L as a mutation-independent global splicing driver in MDS via reduced WTAP; epigenomic remodeling of unmethylated CpGs is an unexpected mechanistic insight; nanoparticle-ASO rescue is a compelling translational proof of concept |
| Clinical Relevance | 4 | Primary human MDS samples used for ex vivo ASO correction is a meaningful translational step, but this remains preclinical — no in vivo human data; cap applied for mixed-species study |
| Population Reach | 6 | MDS affects ~60,000–170,000 people in the US (underdiagnosed); the mutation-independence of MBD1-L means it could apply across all MDS subtypes, which broadens the potential reach considerably |
| Implementation Speed | 3 | ASO/nanoparticle platform requires IND, Phase I–III trials; realistically 8–12+ years to clinical adoption if development proceeds optimally |
| Evidence Strength | 6 | Strong mechanistic data using primary human samples and xenotransplantation; published in Blood; limited by ex vivo/mouse model scope and abstract-only access |
Key quantitative result: Not specified in abstract; functional rescue of erythroid differentiation in primary human MDS samples by nanoparticle-ASO correction.
External validation: Xenotransplantation models provide in vivo proof-of-concept; no independent replication published yet.
Main limitation: No in vivo human or clinical data; xenotransplantation models do not fully recapitulate human MDS biology; abstract-only access limits full methodological assessment.
Equity implications: If developed, a mutation-independent therapy would theoretically benefit MDS patients regardless of their specific mutational profile — reducing the disparity between patients who do/don't have "actionable" mutations. However, ASO/nanoparticle therapeutics are typically expensive and require specialized infusion infrastructure.
Evidence Maturity: ✅ Confirmed Exploratory — compelling mechanistic data, significant translational gap remains.
Article 4 — TIRADS-Based Selective FNA RCT for Thyroid Nodules (PMID 42322189)
| Dimension | Score | Rationale |
|---|---|---|
| Scientific Novelty | 6 | First RCT design for EU-TIRADS FNA selection is genuinely novel in the sense of study design; the concept of risk-stratified biopsy is not new, but prospective randomized evidence is rare in this field |
| Clinical Relevance | 7 | Thyroid nodule evaluation is one of the highest-volume diagnostic procedures in endocrinology; improving FNA yield while reducing unnecessary procedures has direct patient and system impact |
| Population Reach | 7 | Thyroid nodules affect ~65% of the general population on ultrasound screening; millions of FNA procedures performed annually worldwide |
| Implementation Speed | 6 | EU-TIRADS is already widely used in Europe; this RCT would support protocol revision, but the limited FNA omission rate (7%) may slow institutional enthusiasm for workflow change |
| Evidence Strength | 6 | RCT design is a strength; n=195 across 4 centers is modest for practice-changing conclusions; unblinded design is a limitation; single-country (Sweden) limits generalizability |
Key quantitative result: Bethesda IV-VI yield: 26% (selective) vs. 13% (non-selective), p=0.039; malignancy rates similar (8% vs. 5%); FNA omission achieved in only ~7% of patients.
External validation: Single-country RCT; no external replication yet.
Main limitation: Small sample size (n=195); unblinded; conducted in a single healthcare system (Western Sweden); FNA omission rate far below retrospective estimates, raising questions about real-world applicability.
Equity implications: Benefits patients who currently undergo unnecessary FNA (avoiding procedure anxiety, cost, complications); EU-TIRADS applicability in lower-resource settings without high-quality ultrasound may be limited.
Evidence Maturity: ✅ Confirmed Validated (with qualification — validated as proof-of-concept RCT, but insufficient alone for practice change).
Article 5 — NTSR1 Defines Immune-Cold CRC Subtype (PMID 42322125)
| Dimension | Score | Rationale |
|---|---|---|
| Scientific Novelty | 7 | NTSR1 as a driver of immune exclusion via PLXNB3/FLNC/AHNAK2 axis is novel; addressing immune-cold MSS-CRC — a major unmet need — with a pre-existing pharmacological antagonist is a high-value translational finding |
| Clinical Relevance | 4 | Two clinical cohorts (TCGA + surgical validation) provide prognostic credibility, but T-cell reversal data are 3D spheroid only; cap applied for mixed-species/in vitro study |
| Population Reach | 7 | MSS-CRC represents |
| Implementation Speed | 4 | SR48692 is a research probe, not a clinical-stage compound; IND filing, Phase I–III development needed; 6–10 year realistic timeline |
| Evidence Strength | 6 | Dual-cohort retrospective analysis with Random Forest ML is methodologically credible for prognostic claim; 3D spheroid reversal is mechanistically informative but not in vivo; published in Annals of Surgery |
Key quantitative result: NTSR1 is independent predictor of worse PFS and OS (specific HRs not available from abstract); functional T-cell exclusion reversal demonstrated in 3D spheroid models with SR48692.
External validation: TCGA discovery + independent surgical cohort validation is a meaningful two-cohort design; no in vivo immune model yet.
Main limitation: In vitro (3D spheroid) immune reversal data cannot confirm in vivo tumor microenvironment dynamics; no in vivo co-culture or syngeneic mouse immune model shown.
Equity implications: If NTSR1 antagonism enables checkpoint immunotherapy in MSS-CRC, the benefit would extend to the vast majority of CRC patients globally currently excluded from immunotherapy — a substantial equity gain. However, combination therapy development will likely be expensive and concentrated in high-income settings initially.
Evidence Maturity: ✅ Confirmed Exploratory — novel target with clinical prognostic support, requires in vivo immune validation.
Article 6 — MPQC Framework for Clinical ML Quality Control (PMID 42321983)
| Dimension | Score | Rationale |
|---|---|---|
| Scientific Novelty | 6 | The problem (ML QC in clinical settings) is widely recognized; the MPQC scenario-based framework is a concrete and practical proposed solution — more systematized than prior approaches, though concept is incremental |
| Clinical Relevance | 6 | Addresses a genuine patient safety gap as ML algorithms proliferate in clinical labs; not a clinical treatment advance but a governance/safety innovation with real downstream patient impact |
| Population Reach | 5 | Relevant wherever clinical ML algorithms are deployed — potentially broad, but currently limited to institutions already using such algorithms |
| Implementation Speed | 6 | Framework is conceptually simple and analogous to existing lab QC practices; adoption depends on regulatory pressure and institutional governance, not new technology |
| Evidence Strength | 4 | Perspective/methodological paper with single illustrative case study (mSTOP); no large-scale prospective validation of MPQC performance published |
Key quantitative result: No primary quantitative results; illustrative case study only.
External validation: None published; conceptual framework only.
Main limitation: No empirical validation that MPQC as implemented detects clinically meaningful algorithmic failures at acceptable sensitivity/specificity; single illustrative example.
Equity implications: If widely adopted, MPQC would protect all patients equally from algorithmic failures; however, resource-intensive implementation may favor well-resourced academic medical centers, potentially widening the quality gap with community settings.
Evidence Maturity: ✅ Confirmed Exploratory — practical framework proposal, validation pending.
Article 7 — POR-BG@Alb Nanosensitizer for Sonodynamic Therapy (PMID 42322055)
| Dimension | Score | Rationale |
|---|---|---|
| Scientific Novelty | 7 | Identifying SDT-induced PD-L1/CD47 upregulation as an immune resistance mechanism and engineering a nanosensitizer to block it simultaneously is a genuinely novel dual-function concept |
| Clinical Relevance | 3 | Animal study only; cap at 5 applied per non-human study rule — further reduced to 3 given no clinical translation data and significant engineering-to-clinical gap |
| Population Reach | 5 | Bladder and breast cancer models; if validated in humans, would have broad relevance, but this is speculative at current stage |
| Implementation Speed | 2 | Novel nanoparticle formulation requires extensive CMC development, toxicology, Phase I–III; 10+ year realistic timeline |
| Evidence Strength | 4 | Three mouse cancer models provides some consistency; no human data; Advanced Materials publication suggests methodological rigor in materials science, but clinical translation rigor is not assessed |
Key quantitative result: Median survival extended from 17 to 44 days in xenograft mouse model.
External validation: Three independent mouse cancer models is internally consistent; no independent replication.
Main limitation: Mouse-only data; xenograft models do not fully recapitulate human immune biology; nanoparticle pharmacokinetics in humans are unknown.
Equity implications: Premature to assess; sonodynamic therapy requires specialized ultrasound equipment which may limit access in lower-resource settings.
Evidence Maturity: ✅ Confirmed Exploratory — interesting preclinical proof-of-concept, large translational gap.
Article 8 — AI Digital Morphology for Blast Preclassification: Narrative Review (PMID 42322062)
| Dimension | Score | Rationale |
|---|---|---|
| Scientific Novelty | 4 | The limitations of AI in bone marrow aspirate analysis are known; this review synthesizes current knowledge without new data; ICSH position is useful but not novel |
| Clinical Relevance | 5 | Directly relevant to hematology lab practice; the "not ready for standalone use" conclusion is clinically important to prevent premature adoption |
| Population Reach | 5 | Relevant to all institutions performing bone marrow analysis for AML/MDS diagnosis; hematopathology labs globally |
| Implementation Speed | 4 | The conclusion actually slows implementation — identifies gaps that need filling before adoption |
| Evidence Strength | 4 | Narrative review without systematic methodology; no original data; Int J Lab Hematol is an appropriate venue |
Key quantitative result: >90% sensitivity/specificity for peripheral blood smear blast detection; bone marrow aspirate performance "variable and insufficient" (specific metrics not provided in abstract).
External validation: Synthesizes multiple published studies; no primary data or meta-analytic pooling.
Main limitation: Narrative (not systematic) review methodology; selection bias in included studies cannot be excluded.
Equity implications: AI-assisted morphology could democratize expert hematopathology review in resource-limited settings if validated — but current evidence supports caution before deployment.
Evidence Maturity: ✅ Confirmed Exploratory — field synthesis, not primary evidence.
Article 9 — Engineered Neutrophils Review (PMID 42322170)
| Dimension | Score | Rationale |
|---|---|---|
| Scientific Novelty | 5 | Reviews an emerging convergence field (CRISPR + nano + AI in neutrophils); no new findings |
| Clinical Relevance | 3 | No human data; distant from clinical application |
| Population Reach | 4 | Broad potential if engineered neutrophils succeed, but this is speculative |
| Implementation Speed | 2 | Multiple overlapping translational barriers; 10+ years realistically |
| Evidence Strength | 3 | Narrative review, no original data, unknown species model |
Evidence Maturity: ✅ Confirmed Exploratory — landscape review only.
Article 10 — Inflammation and Fatigue in CRC Survivors (PMID 42322026)
| Dimension | Score | Rationale |
|---|---|---|
| Scientific Novelty | 4 | Inflammation-fatigue link in cancer survivors is an active but not novel area; longitudinal design adds value over cross-sectional studies |
| Clinical Relevance | 4 | Cancer-related fatigue is undertreated; if specific inflammatory biomarkers predict fatigue trajectory, that could guide supportive care — but full results not available |
| Population Reach | 6 | CRC survivors number in the millions globally; fatigue affects ~30–60% post-treatment |
| Implementation Speed | 4 | Supportive care interventions are rapidly implementable in principle, but the specific biomarker-to-intervention pathway is not yet defined |
| Evidence Strength | 3 | Low confidence classification; title-only access; longitudinal observational design has confounding risk |
Evidence Maturity: Confirmed Exploratory — insufficient data available to confirm maturity; downgrade from classification_confidence = low.
Article 11 — Gene Delivery for Cerebral Neurodegeneration Review (PMID 42322051)
| Dimension | Score | Rationale |
|---|---|---|
| Scientific Novelty | 4 | AAV/ASO/nanoparticle landscape for CNS is well-documented; emerging approaches (engineered capsids, FUS-mediated BBB opening) add some novelty |
| Clinical Relevance | 4 | Relevant for neuro gene therapy clinicians, but not core watchlist scope; no primary data |
| Population Reach | 5 | Neurodegenerative diseases affect tens of millions globally; review scope is broad |
| Implementation Speed | 2 | CNS gene therapy is at early clinical stage for most conditions |
| Evidence Strength | 3 | Narrative review; industry-funded (Ocugen); unknown species model; medium classification confidence |
Note: Industry-funded review (Ocugen) — potential framing bias should be considered.
Evidence Maturity: ✅ Confirmed Exploratory — review only, peripheral watchlist match.
PHASE 3 — Ranking
Conflict Check
There are no directly conflicting findings across articles in this batch. Articles 1, 2, and 4 all support more precise patient selection in their respective domains (CLL, CRC, thyroid). Articles 3 and 5 address complementary aspects of CRC biology (MDS splicing vs. CRC immune exclusion) with no contradictions. The AI morphology review (Article 8) and MPQC framework (Article 6) address complementary AI governance problems without conflict.
Composite Impact Scores
Weighting: Clinical Relevance 30% | Population Reach 25% | Scientific Novelty 20% | Implementation Speed 15% | Evidence Strength 10%
| Rank | Article | Title (linked) | Impact Score | Clin. Rel. (×0.30) | Pop. Reach (×0.25) | Sci. Nov. (×0.20) | Impl. Speed (×0.15) | Evid. Str. (×0.10) | OpenClaw Triage | Study Design | Priority Flag |
|---|---|---|---|---|---|---|---|---|---|---|---|
| #1 | 2 | Tumour Budding Meta-Analysis, CRC | 7.45 | 8 | 8 | 6 | 7 | 8 | 8 | SR/Meta-analysis (43 studies, n=17,831) | 🟢 Near-term implementable |
| #2 | 1 | Ibrutinib Genetic Substudy, CLL12 | 7.35 | 8 | 6 | 7 | 8 | 8 | 8 | Retrospective genetic substudy of RCT (n=515) | 🟢 Near-term implementable |
| #3 | 4 | TIRADS RCT, Thyroid FNA | 6.55 | 7 | 7 | 6 | 6 | 6 | 7 | Multi-centre RCT (n=195) | 🟢 Near-term implementable |
| #4 | 3 | MBD1 Splicing in MDS | 5.60 | 4 | 6 | 9 | 3 | 6 | 7 | Mechanistic preclinical + primary human samples | ⚪ Promising but preliminary |
| #5 | 5 | NTSR1 Immune-Cold CRC | 5.45 | 4 | 7 | 7 | 4 | 6 | 7 | Dual-cohort retrospective + in vitro | ⚪ Promising but preliminary |
| #6 | 6 | MPQC Framework for ML QC | 5.40 | 6 | 5 | 6 | 6 | 4 | 5 | Perspective/methodological + case study | 🟢 Near-term implementable |
| #7 | 7 | POR-BG@Alb Nanosensitizer SDT | 4.00 | 3 | 5 | 7 | 2 | 4 | 5 | Preclinical mouse models | ⚪ Promising but preliminary |
| #8 | 8 | AI Digital Morphology Review | 4.55 | 5 | 5 | 4 | 4 | 4 | 4 | Narrative review | ⬜ Standard |
| #9 | 10 | Inflammation & Fatigue in CRC Survivors | 4.10 | 4 | 6 | 4 | 4 | 3 | 4 | Longitudinal observational | ⬜ Standard |
| #10 | 11 | Gene Delivery CNS Review | 3.65 | 4 | 5 | 4 | 2 | 3 | 4 | Narrative review | ⬜ Standard |
| #11 | 9 | Engineered Neutrophils Review | 3.50 | 3 | 4 | 5 | 2 | 3 | 4 | Narrative review | ⬜ Standard |
Note: Article 8 (Digital Morphology) scores marginally above Articles 9–11 on Clinical Relevance given its direct hematopathology practice implications, despite being a review.
Rank Justifications
#1 — Tumour Budding Meta-Analysis 🟢 The largest meta-analysis of tumour budding as a prognostic biomarker in CRC (n=17,831; 43 studies) delivers hazard ratios of 2.4–3.4 across disease-free, overall, and cancer-specific survival in stage II-III disease. These are large, consistent effects across an enormous and globally relevant patient population. Critically, tumour budding assessment requires only standard H&E staining — no new technology, no added cost — meaning the barrier to implementation is primarily pathologist training and workflow standardization, both achievable in the near term. The findings directly support incorporating budding grade into adjuvant chemotherapy decision-making for stage II CRC, one of the most genuinely difficult decisions in GI oncology. While the meta-analysis does not include a therapeutic RCT, the strength and consistency of the prognostic signal across 43 independent studies constitutes compelling evidence for routine adoption.
Why it matters: Hundreds of thousands of stage II CRC patients annually face an uncertain adjuvant chemotherapy decision. A validated, low-cost histological marker that independently predicts mortality risk could spare some patients unnecessary chemotherapy toxicity and ensure others receive treatment they urgently need.
#2 — Ibrutinib CLL12 Genetic Substudy 🟢 With 515 patients and 69 months of follow-up from the rigorous CLL12 RCT, this genetic substudy from the German CLL Study Group delivers precision oncology-grade guidance on ibrutinib in early-stage CLL. The critical finding — that del(17p)/TP53-mutated patients derive no EFS benefit from ibrutinib, and no subgroup gains an OS advantage — immediately informs clinical practice. IGHV, FISH, and TP53 testing are already embedded in CLL workup, making this directly implementable without new diagnostic infrastructure. The findings also sharpen clinical trial design for future early-stage CLL studies and provide data for informed patient counselling about the risks and benefits of early intervention.
Why it matters: CLL is the most common adult leukemia in Western countries. Knowing that ibrutinib fails the highest-risk genetic subgroup (del17p/TP53) in the early-stage setting — rather than potentially helping them — prevents premature treatment with a drug that carries meaningful toxicity and cost burdens.
#3 — TIRADS RCT, Thyroid FNA 🟢 As the first randomized trial of EU-TIRADS for thyroid nodule FNA selection, this study from four Swedish hospitals fills a critical evidence gap in a very high-volume clinical workflow. The doubling of high-risk cytology yield (26% vs. 13%) without missing cancers is an encouraging proof-of-concept, but the modest sample size (n=195) and surprisingly low FNA omission rate (~7%) temper enthusiasm for immediate protocol adoption. The study is valuable for informing next-generation TIRADS validation trials and adds credibility to risk-stratified biopsy selection, but further larger multicenter replication is needed before major guideline revision.
Why it matters: Thyroid nodules are found in the majority of adults on ultrasound; reducing unnecessary FNA procedures while maintaining cancer detection accuracy could spare millions of patients annually from an anxious, uncomfortable, and sometimes inconclusive procedure.
#4 — MBD1 Splicing Driver in MDS ⚪ The discovery of MBD1-L as a mutation-independent, globally operative splice variant in MDS — driven by reduced WTAP expression and reversible by nanoparticle-ASO correction in primary human samples — is the most scientifically novel finding in this batch. Its applicability across MDS subtypes regardless of SF3B1/SRSF2 mutational status is especially significant, because it creates a therapeutic target for patients currently without molecularly targeted options. The translational gap is substantial (preclinical/ex vivo only), but the combination of mechanistic depth, primary human sample validation, and a clear therapeutic modality (ASO) in a disease with genuine unmet need makes this a strong watchlist entry.
Why it matters: MDS remains a disease with limited treatment options for most patients; a mutation-agnostic RNA therapeutic target could eventually benefit the entire MDS population rather than a small molecularly defined subset.
#5 — NTSR1 Immune-Cold CRC ⚪ NTSR1 joins a growing list of proposed immune exclusion targets in MSS-CRC, but the combination of clinical prognostic validation (two cohorts), mechanistic pathway definition (PLXNB3/FLNC/AHNAK2), and functional reversal with a pre-existing pharmacological probe (SR48692) distinguishes this from pure biomarker discovery. The 3D spheroid immune reversal is conceptually compelling but limited. The enormous unmet need — approximately 1.4 million immune-cold CRC patients annually who do not benefit from checkpoint blockade — justifies close tracking of this target.
Why it matters: If NTSR1 antagonism can genuinely convert immune-cold CRC to immune-responsive tumors, it could unlock checkpoint immunotherapy for the overwhelming majority of CRC patients currently excluded from it.
#6 — MPQC Framework for Clinical ML 🟢 A practical, implementable framework for monitoring deployed ML algorithms in clinical labs addresses an increasingly urgent patient safety gap. The MPQC approach is conceptually sound and analogous to familiar laboratory QC paradigms, which supports rapid adoption. Its current limitation is the absence of empirical validation data — it is an expert proposal illustrated by a single case study. As regulatory agencies begin requiring prospective monitoring of deployed clinical AI, MPQC-type frameworks will move from optional to required, making this paper timely for lab directors and clinical informaticists.
#7–#11 — Lower-ranked articles are either animal-only preclinical studies, narrative reviews without original data, or articles with low confidence classification. They contribute useful field context but do not meet the threshold for near-term clinical impact.