📅 PUBMED DAILY TRIAGE — Sunday, June 7, 2026

Window: 2026-06-06 09:00 UTC → 2026-06-07 09:00 UTC (CRDT) Topics scanned: 10 | Candidates reviewed: ~90 unique PMIDs | Kept: 20 High priority: 6 | Standard: 14 | Low/archive: ~28 | Unsolicited finds: 1 Run status: PARTIAL_SUCCESS ⚠️ (T3 liquid biopsy abstracts not retrieved — 3 PMIDs deferred)

🔴🟠 HIGH PRIORITY (8+) — Best Finds by Topic

Cardiovascular-Metabolic

TR-2026-06-07-001 · CHG Index + CKM Syndrome continuum (UK Biobank n=370,916 + Beijing Anzhen n=8,494) · 9/10 Association of CHG composite index (Cholesterol, HDL-C, Glucose) with CKM syndrome onset, stage-wise progression, and MACCE across two large prospective cohorts (combined n=379K, up to 16.5yr follow-up); causal forest ML identifies high-HbA1c/low-inflammation subgroup as highest-impact target. First study to track CKM continuum with a simple 3-parameter index from routine labs. (Diabetol Metab Syndr; DOI: 10.1186/s13098-026-02192-2)

AI/ML Diagnostics + Rare Diseases (crossover)

TR-2026-06-07-002 · G.AI platform for rare disease variant interpretation (39,156 multicenter WES) · 9/10 Multicenter validation of AI platform for rare disease genomic diagnosis: 94% HPO phenotype concordance, 95% Top-1 / 99.6% Top-20 variant accuracy, analysis time reduced from 4-6h → 48min across 39,156 cases. Directly addresses diagnostic scalability bottleneck in rare disease genomics; 100% Top-3 accuracy for metabolic disorders. Conflict of interest noted (corresponding author = Xbiolabs founder). (J Transl Med; DOI: 10.1186/s12967-026-08368-8)

Hematologic Malignancies + Novel Therapeutics (crossover)

TR-2026-06-07-003 · Fludarabine dose escalation does NOT improve CAR-T outcomes — EBMT registry (n=1,498) · 8/10 Largest registry analysis of lymphodepletion dosing in CAR-T for LBCL: dose escalation (82.6–120 mg/m²) associated with inferior OS vs. standard dose (HR 1.29, p=0.036) in tisa-cel; axi-cel at standard dose outperformed both tisa-cel dose groups (PFS/OS); higher fludarabine does not reduce relapse. Directly practice-changing for lymphodepletion conditioning protocols. (Bone Marrow Transplant; DOI: 10.1038/s41409-026-02925-x)

AI/ML + Aging (crossover)

TR-2026-06-07-004 · Proteomic aging clocks + deep learning for eye disease prediction (>55,000 transethnic participants) · 8/10 Cross-national study integrating high-throughput proteomics with DL retinal phenotyping validates proteomic aging acceleration as a cross-disease biomarker for cataract, diabetic retinopathy, AMD, and glaucoma — adaptable across sexes and ethnicities. Streamlined cost-effective clock proposed. Positions aging proteomics + AI as scalable eye disease surveillance platform. (npj Digit Med; DOI: 10.1038/s41746-026-02805-0)

Aging/Population Health

TR-2026-06-07-005 · Frailty — not sex — drives ICU mortality in elderly: multinational pooled analysis (n=10,363) · 8/10 Pooled analysis of VIP1+VIP2+COVIP prospective registries (n=10,363, age ≥70-80): pre-admission frailty increases adjusted mortality 8% per CFS point; after frailty adjustment, male sex has only small residual excess mortality (IRR 1.08); Bayesian analysis shows >10% sex-based mortality difference is unlikely (probability 2-13%). Supports frailty-first ICU triage over sex-based stratification; sex-specific frailty prevalence (women 35.3% vs men 25.6%) highlights need for frailty-sensitive care in older women. (Eur J Intern Med; DOI: 10.1016/j.ejim.2026.106986)

💡 UNSOLICITED FIND

TR-2026-06-07-006 · Antihypertensive treatment + breast cancer survival in Black women (n=2,474) · 8/10 Prospective cohort (Black Women's Health Study): treated hypertension strongly inversely associated with breast cancer-specific death in ER+ cases (HR 0.53, 95% CI 0.34-0.83); no signal in full cohort (HR 0.81, non-significant); opposite trend in ER- cases. Equity signal: Black women face highest breast cancer mortality burden and highest hypertension rates. Potential drug-repurposing hypothesis but observational design limits causal claims. unsolicited_find = true; from sentinel scan. (Breast Cancer Res; DOI: 10.1186/s13058-026-02317-5)

🟢⚪ STANDARD (5–7) — Top by Topic

Hematologic Malignancies

• PMID 42251161 · BEND4 AML prognostic biomarker · 7/10 · Genomics/Precision Medicine — Novel single-gene RT-qPCR marker for adverse AML (91% sensitivity, 81% specificity), validated in n=1,688; mixed human/preclinical
• PMID 42251370 · Baseline immune composition + CAR-T expansion in R/R LBCL (n=33) · 6/10 · Treatment Innovation — Exploratory; CD4/CD8 ratio associated with longer PFS (HR 0.41); n=33, hypothesis-generating
• PMID 42251338 · TNFRSF/Treg roadmap for GvHD · 6/10 · Treatment Innovation — Comprehensive review; TNFR2 and DR3 agonism as selective Treg expansion strategies preserving GvL

CBC/ML in Hematology

• No high-priority finds today (0 results this window — consistent with prior runs)

Liquid Biopsy / ctDNA

• No high-priority finds today (3 PMIDs retrieved — 42250735, 42250371, 42250175 — but abstracts not fetched due to batch truncation; deferred to next run)

AI/ML in Diagnostics

• PMID 42251454 · [18F]PSMA-3Q PET/CT radiomics for prostate cancer grading · 7/10 · Diagnostics — AUC 0.917 for ISUP ≥4 prediction; n=243; pending external validation
• PMID 42251145 · AI foundation models for EC molecular subtyping — real-world validation · 7/10 · Diagnostics — UNI model AUROC 0.844 for p53abn subtype; real-world heterogeneous data; scanner-robust
• PMID 42251301 · AI in prehospital ACS assessment — scoping review (n=319,709) · 6/10 · Diagnostics — AUC 0.81-0.99 for AI-based ECG prehospital diagnosis; maps evidence gaps
• PMID 42251432 · scRNA + ML in HPV+ cervical cancer LLPS biomarkers · 6/10 · Diagnostics — 70,519 cells; RPL5/RPL11 suppression as HPV+ cancer signature; 69.7% classification accuracy

Precision Oncology

• PMID 42251161 (see Hematologic above — T1/T5 crossover)

Novel Therapeutics

• PMID 42250977 · T2 endotypes in airway diseases (asthma + COPD) — review · 5/10 · Treatment Innovation — T2-COPD affects 20-30% of COPD; 2 biologics now approved; narrative review

Cardiovascular-Metabolic

• PMID 42251396 · Aerobic training body composition meta-analysis (postmenopausal women, 16 RCTs, n=1,571) · 7/10 · Prevention — Fat mass -1.83 kg, waist circumference -2.02 cm; lean mass uncertain; no clear dose-response
• PMID 42251331 · AIPFI + cardiometabolic multimorbidity longitudinal (CHARLS n=7,995, 9yr) · 7/10 · Prevention — Composite atherogenic×frailty index, Q4 HR 2.46 (CI 1.77-3.43) for incident CMM; SHAP model
• PMID 42251267 · TyG-ABSI COPD predictor, cross-population (CHARLS + NHANES, XGBoost AUC >0.96) · 6/10 · Diagnostics — Metabolic composite index validated in China and US populations
• PMID 42251129 · Frailty + sex differences in BP control, older adults Vietnam (n=1,038) · 5/10 · Public Health — Frailty-BP association female-specific (OR 2.01); cross-sectional

Aging/Population Health

• PMID 42250989 (HIGH — see above)
• PMID 42251449 · Frailty indices in spinal metastasis surgery SR/MA (12 studies, n=17,446) · 6/10 · Diagnostics — mFI-11 OR 2.94 for complications; JHACG predicts discharge/LOS; survival inconsistent

Rare Diseases

• PMID 42251412 (HIGH — G.AI platform, see above)
• PMID 42251206 · Early onset appendiceal adenocarcinoma (EOAA, n=49 vs n=132 AOAA) · 6/10 · Other — More aggressive clinical course, late recurrences in 60% of 5yr survivors; rising incidence

📊 TODAY'S BREAKDOWN

By research category:

• Diagnostics: 9
• Treatment Innovation: 5
• Prevention: 3
• Genomics/Precision Medicine: 2
• Public Health: 1

By evidence maturity:

• Exploratory: 13
• Validated: 5
• Potentially Practice-Changing: 2 (TR-001 CHG Index, TR-003 CAR-T Fludarabine)

By watchlist topic (reviewed / kept / highest score):

By source type: peer_reviewed 20 | preprint 0

Topics with zero 8+ articles this run: CBC/ML, Liquid Biopsy (T3 partial failure), Precision Oncology (standalone)

🔧 RUN NOTES

1. T3 PARTIAL FAILURE: 3 liquid biopsy PMIDs (42250735, 42250371, 42250175) retrieved via esearch but abstracts not fetched due to API batch truncation. No window expansion triggered (>5 relevant records found across other topics). Recommend adding these 3 PMIDs to deferred_pmids_next_run.

2. T2 (CBC/ML): 0 results — 3rd consecutive day without results. Query broadening recommended: add cytology[tiab] OR morphology[tiab] OR differential[tiab] OR peripheral smear[tiab] as synonyms.

3. T8 animal filter: NOT animals[MeSH] + human keyword requirement working well; one slipthrough (PMID 42251437, J Anim Sci Biotechnol) — animal journal content passed due to "human" synonym in abstract. Suggest adding NOT (J Anim Sci[jour] OR Poultry Sci[jour] OR Livestock Sci[jour]) to T8 query.

4. T6 clinical filter (AND (clinical[tiab] OR trial[tiab] OR patient[tiab] OR randomized[tiab])): Confirmed working — signal-to-noise improved, no purely preclinical noise in top 10. Keep for future runs.

5. Sentinel (watchlist-anchored): yielded 1 strong unsolicited find (TR-006, PMID 42251447, breast cancer + antihypertensive treatment in Black women). Anchor approach confirmed effective. Keep query.

6. TOP FINDING TODAY: TR-2026-06-07-001 (CHG Index + CKM continuum, score 9/10) and TR-2026-06-07-002 (G.AI rare disease genomics platform, score 9/10) tie for highest signal.

7. T5 partial: 2 records (42250592, 42250565) not retrieved due to batch limits; remaining records were predominantly preclinical mechanism studies unlikely to change priority distribution.