Skip to content

March 2025 — Monthly Briefing

1. What was read

812 research articles from 33 of the 33 journals on the roster.

A further 275 records (25% of the 1,087 this month's harvest returned) carry no abstract and are not research: news, correspondence, editor highlights and obituaries. PubMed types them Journal Article and nothing else, so the publication-type filter cannot see them — across the whole index they are 22.7% of it and 35% of general, and 3,801 of them are JAMA. Every count below is over the full harvest and is inflated by this; the figure above is not.

General Cancer Immune Bioinfo Sysbio
486 200 88 287 26

Largest contributors: JAMA 118, Nature 106, Bioinformatics 102, Science 79, Nucleic Acids Res 68, Lancet 52.

2. Net change on the thread curve

Thread This month Cumulative Status
agentic-ai — +9 / −1 forming
cell-therapy +7 +160 / −47 forming
ctdna-mrd — +48 / −22 forming
foundation-models +3 / −3 +87 / −59 splitting
immunotherapy-resistance +4 / −1 +173 / −64 forming
interception +1 / −1 +92 / −21 forming
method-credibility −18 +12 / −360 crisis
neoantigen-vaccines +4 / −3 +62 / −28 forming
neuro-immune-microbiome-axis +7 +129 / −14 forming
next-gen-checkpoints +2 / −1 +54 / −35 splitting
single-cell-3d-genome +3 +54 / −17 forming
suppressive-population-inversion +1 / −6 +38 / −61 contested

Status is derived by rule, not asserted. Fewer than 5 cards → too-early; then refutes ≥ 2×supports → crisis; refutes > supports → contested; supports ≥ 2×refutes → forming; else splitting.

3. The evidence that moved it

The templated preclinical block hits 33, its highest strict count in the corpus. Thirteen in J Immunother Cancer, six in Cancer Cell, five in Nature Cancer. This is the month to hold in mind whenever a later briefing claims the genre is declining: it is not the start of a decline, it is the peak of a series that turns out to have no trend at all once every month is counted by the same rule.

Neoantigen vaccines split inside one journal issue. 39972124 reports that RNA neoantigen vaccines prime long-lived CD8 T cells in pancreatic cancer; 39972134 reports that clonal driver neoantigens are lost under TKI and immune selection pressure. Same journal, same date. The split is not between labs — it is between the clinical arm and the target-space arm.

Corrected 2026-09-04

This paragraph said the thread "has been splitting from the month it opened". It has not. The 2026-09-04 abstract re-scan took neoantigen-vaccines from 52 cards to 151, and on the completed record it reads forming in 32 of its 44 months, this one included; the old curve had it splitting in 19. The split described above is real and visible in this month's two papers — it is the shape of the argument, not the shape of the ratio, and the ratio never supported the stronger claim.

Low-dose gut irradiation improves PD-L1 blockade in patients (40068595, Cancer Cell). Worth flagging because it is one of very few microbiome-adjacent interventions in the corpus with a human readout rather than a mouse one, in a thread whose weakness is precisely that its microbial arm never reaches patients.

4. Refutations

(a) Landing on a tracked thread

40023157 (Immunity): IL-17RA signalling provides a dual tumour-suppressor function in late-stage colorectal carcinogenesis. IL-17 is targeted therapeutically on the assumption that it is inflammatory and pro-tumour.

39551185 (Ann Oncol): tumour and blood B-cell abundance outperforms established checkpoint-response prediction signatures in head and neck cancer. A simple cell-count beating the signature layer is the same shape as the computational refutations that arrive in 2026, arriving here from the wet side.

(b) Method credibility

Eight cards, and unusually they cluster on single-cell rather than on genomics. 40069819: single-cell RNA-seq data carry prevalent blood contamination. 40098192: single-cell differential expression has shortcomings requiring a new statistical paradigm. 40234247: single-cell classifiers are vulnerable to adversarial attacks. Three independent statements, in one month, that the most-used analysis stack in the corpus is unreliable at different layers — contamination in the input, miscalibration in the test, brittleness in the classifier.

39965933: the transcriptomic and proteomic effects of gene deletion are not evolutionarily conserved. Every cross-species inference from a knockout inherits this.

40119729: spuriously transcribed RNAs from CRISPR-sgRNA expression plasmids scaffold condensates and hamper accurate genomic imaging — an artifact created by the tool, visible only to people looking for it.

39891002 (Nat Med): familial confounding underlies the reported associations between maternal health and autism.

40164443: retraction and replacement of a histone methyltransferase paper — the rarer and more honourable form of correction, where the replacement runs in the same issue.

5. Open items

Nothing carried forward.

6. Meetings

No abstract book published this month. The abstract layer runs 2023-02 to 2026-08 and is seasonal by nature — meetings are annual, so most months are empty here and that is the expected shape, not a gap in the harvest.

7. Regulatory

Event Count
trial-first-posted 517 (of which 458 genuinely oncology)
ema-authorisation 6
fda-approval 4

4 FDA actions — dated from each notice's own prose, not the table's posting date, which disagrees on 7 of 190 rows and once by enough to move an approval into the wrong month:

  • 2025-03-19 — FDA approves pembrolizumab for HER2 positive gastric or gastroesophageal junction adenocarcinoma expressing PD-L1 (CPS ≥1)
  • 2025-03-26 — FDA approves cabozantinib for adults and pediatric patients 12 years of age and older with pNET and epNET
  • 2025-03-28 — FDA approves durvalumab for muscle invasive bladder cancer
  • 2025-03-28 — FDA expands Pluvicto’s metastatic castration-resistant prostate cancer indication

3 EMA oncology authorisations:

  • 2025-03-07 — Rytelo
  • 2025-03-28 — Dyrupeg
  • 2025-03-28 — Tivdak

8. Trade press

40 on-topic of 341 items across 3 feeds (n-stat 330, n-aacr-blog 8, n-acir 3). Headlines are selected by subject keyword, not recency — most of what these feeds carry is health policy and industry finance the corpus does not read.

  • 2025-03-28 — Editors’ Picks, March 2025: Cancer Cell Senescence, Ultraviolet Protection, and More — n-aacr-blog
  • 2025-03-27 — Personalized Vaccines and Beyond: Confronting the Complex Biology of Kidney Cancer — n-aacr-blog
  • 2025-03-27 — STAT+: Two of FDA’s top cancer regulators to depart, heightening worries about drug reviews — n-stat
  • 2025-03-27 — More HHS upheaval, 23andMe’s bankruptcy, and AI for cancer survival — n-stat
  • 2025-03-27 — STAT+: New tests for Alzheimer’s and cancer get FDA’s ‘breakthrough’ status — n-stat
  • 2025-03-26 — STAT+: Altis says its AI tool can cut risk in cancer trials — n-stat
  • 2025-03-26 — STAT+: Wave to submit Duchenne drug for approval after Phase 2 success — n-stat
  • 2025-03-26 — STAT+: This Canadian company’s AI tool could help drugmakers cut risk in late-stage cancer trials — n-stat
  • 2025-03-26 — From TCRs to immunotherapy targets: detection of antigens in pediatric brain tumors — n-acir
  • 2025-03-25 — Opinion: STAT+: Impatient about cell and gene therapy? Progress in biotech is not always linear — n-stat
  • 2025-03-24 — STAT+: Cancer research, long protected, feels ‘devastating’ effects under Trump — n-stat
  • 2025-03-20 — Cancer Today’s Spring 2025 Issue: How Liquid Biopsies Can Guide Treatment Decisions, Extending Survival in Treatment-resistant Breast Cancer, and More — n-aacr-blog
  • … and 28 more on-topic items.