June 2024 — Monthly Briefing¶

1. What was read¶

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

A further 127 records (14% of the 927 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
389	134	84	294	26

Volume swing against the median month: cancer -28%

Check the source before reading anything into this. Two benign causes account for most of these. bioinfo every January carries the Nucleic Acids Research database issue and runs high. And several journals simply deposit nothing in some months: across the 44 months read, 4 of 33 journals are absent from at least one, worst being Nat Mach Intell (20 months missing, 81 papers in total) and Bioinformatics (4 months missing, 2,905 papers in total). Every journal on the roster contributed this month. Queried directly, PubMed returns zero for those journal-months too — the holes are in the source, reproduced exactly by re-running the harvest. So a swing here is not evidence that a field went quiet, and is not usually evidence that the pipeline broke either.

Largest contributors: Bioinformatics 109, Nature 102, Nucleic Acids Res 89, Science 77, Lancet 43, JAMA 38.

2. Net change on the thread curve¶

Thread	This month	Cumulative	Status
`agentic-ai`	—	+4 / −0	too-early
`cell-therapy`	+4 / −1	+108 / −36	forming
`ctdna-mrd`	−2	+34 / −17	forming
`foundation-models`	+4 / −3	+48 / −33	splitting
`immunotherapy-resistance`	+10 / −2	+122 / −39	forming
`interception`	+2 / −2	+69 / −13	forming
`method-credibility`	+1 / −24	+6 / −230	crisis
`neoantigen-vaccines`	—	+49 / −19	forming
`neuro-immune-microbiome-axis`	+10 / −1	+90 / −11	forming
`next-gen-checkpoints`	+1 / −4	+35 / −27	splitting
`single-cell-3d-genome`	+3 / −2	+35 / −15	forming
`suppressive-population-inversion`	+1 / −4	+27 / −40	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¶

Hallucination detection lands in Nature. 38898292: detecting hallucinations in large language models using semantic entropy. The December 2025 briefing calls auditing a generative model's own output — rather than its training or its benchmarks — "a new category". It is in Nature eighteen months earlier. This is the fifth first-appearance claim in this corpus to move earlier on extension, and at this point the individual corrections matter less than the rule: first appearance is a property of the window.

Two acetylcholine papers, same journal, same month. 38744292 and 38744291 (Immunity): tuft-cell acetylcholine released into the gut lumen to promote anti-helminth immunity, and tuft-cell-derived acetylcholine driving epithelial chloride secretion and helminth clearance. With 38772371 (Cell) — innervation of nociceptor neurons in the spleen promoting germinal-centre responses — June 2024 has nerves controlling a lymphoid organ and a named neurotransmitter doing immune work, in the corpus's top venues, in one month.

4. Refutations¶

(a) Landing on a tracked thread¶

Nineteen of this month's refutations land on a tracked thread other than method-credibility. Four groups of them carry the month.

The lead second-generation checkpoint is read out against a control twice, and adds nothing either time. 38723227 (JCO) — RELATIVITY-060 missed its primary endpoint: adding relatlimab (anti-LAG-3) to nivolumab plus chemotherapy in LAG-3-expressing gastric and gastro-oesophageal junction cancer gave a numerically worse response rate (48% against 61%), worse progression-free survival (HR 1.41) and more grade 3–4 toxicity than nivolumab plus chemotherapy alone — in a biomarker-selected population, which is the condition under which this class was supposed to work. 38689060 (Nat Med) — in resectable non-small-cell lung cancer, neoadjuvant nivolumab with relatlimab gave a major pathological response of 30% against 27% for nivolumab alone, with complete resection in 90% against 100%. next-gen-checkpoints has been carrying registry activity as its evidence — whether a nominated target reached a phase 1 — and this is what the readouts look like when they arrive.

The myeloid checkpoint destroys the therapy it was meant to complement. 38750365 (Nature) — combining anti-CD47 antibodies with adoptive T-cell therapy abrogated the benefit outright: macrophages cleared the CAR and TCR T cells fast enough that the antibody worked as a safety switch, and only a CD47 variant invisible to the antibody rescued the combination. It refutes next-gen-checkpoints and cell-therapy at once, and it is a combination failure neither thread could have anticipated from its own cards.

Four cards say the suppressive population was doing something. 38788719 (Cancer Cell) — TREM2, the canonical immunosuppressive myeloid target being blocked in peripheral cancers, is protective in glioblastoma: genetic or pharmacological deficiency accelerated progression, and overexpression impeded growth and synergised with anti-PD-1. 38759656 (Cancer Cell) — effective immunotherapy in two murine models required a late-stage activated M1-like macrophage population recruited via CCR5 by therapy-induced CD8 T cells, and the authors warn explicitly that broad macrophage-targeting combinations would remove it. 38692280 (Immunity) — depleting visceral-adipose Treg cells, or Treg-derived oncostatin-M, or its receptor on stromal cells, strongly impaired insulin sensitivity, on an axis the authors report as conserved in humans. And 38811815 (Nat Immunol) — bone-marrow neutrophils polarised with IL-4 and G-CSF secrete growth factors and drive optic nerve and spinal cord axon regeneration, which is the same argument from the other side: the therapeutic move is reprogramming, not depletion.

A molecular residual-disease endpoint gives the wrong answer inside a randomised trial. 38452207 (Blood) — GRAAPH-2014 de-escalated consolidation in Philadelphia-positive ALL using BCR::ABL1 molecular response as the endpoint. The omission arm met non-inferiority for major molecular response, 77.2% against 71.1%, and then relapsed at 31.3% at four years against 13.2%; the DSMB halted the randomisation. Beside it 38142939 (Ann Oncol) has ctDNA in high-grade osteosarcoma refining a risk score and never guiding therapy, with the assay's sensitivity at surgery and end of treatment too low to follow disease at all. ctdna-mrd closes −2 with nothing on the other side.

Corrected 2026-09-04

This section read, in full: "Nothing this month lands a refutation on a tracked thread, and that is worth recording rather than padding. The month's refutations are all methodological." Nineteen refutations land on a tracked thread other than method-credibility, including two randomised trials of the field's lead second-generation checkpoint. June 2024 had been carded from titles; none of the trial results above announces its outcome in its title — "Neoadjuvant nivolumab with or without relatlimab", "First-Line Nivolumab and Relatlimab Plus Chemotherapy" — so a title-only reading saw a month of benchmarks and concluded the biology was quiet. On next-gen-checkpoints it is in fact the joint-largest month in the corpus to that point — seven cards, tying November 2023 — and four of those seven refute.

(b) Method credibility¶

38886757 (Genome Biol): beyond benchmarking, towards predictive models of dataset-specific pipeline performance. The claim is that a benchmark ranking does not transfer between datasets — which, if true, undercuts the evidentiary value of the benchmark papers that fill this journal every month, including 37537502 in the same month systematically benchmarking single-cell ATAC protocols.

38940140: BioCoder, a benchmark for bioinformatics code generation with LLMs. The corpus's July 2026 briefing treats the arrival of benchmarks as the maturity signal for the agentic thread. Here the benchmark arrives alongside the capability rather than after it, which weakens benchmarks-as-maturity-signal as a general reading.

5. Open items¶

Nothing carried forward.

6. Meetings¶

5,885 abstracts from 1 meeting.

Meeting	Abstracts	With full text
ASCO Annual Meeting 2024	5,885	5,885

7. Regulatory¶

Event	Count
trial-first-posted	495 (of which 449 genuinely oncology)
ema-authorisation	9
fda-approval	8

8 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:

2024-06-06 — FDA approves imetelstat for low- to intermediate-1 risk myelodysplastic syndromes with transfusion-dependent anemia
2024-06-12 — FDA approves selpercatinib for RET fusion-positive thyroid cancer
2024-06-13 — FDA grants accelerated approval to repotrectinib for adult and pediatric patients with NTRK gene fusion-positive solid tumors
2024-06-14 — FDA approves blinatumomab as consolidation for CD19-positive Philadelphia chromosome-negative B-cell precursor acute lymphoblastic leukemia
2024-06-14 — FDA approves durvalumab with chemotherapy for mismatch repair deficient primary advanced or recurrent endometrial cancer
2024-06-17 — FDA approves pembrolizumab with chemotherapy for primary advanced or recurrent endometrial carcinoma
2024-06-21 — FDA grants accelerated approval to adagrasib with cetuximab for KRAS G12C-mutated colorectal cancer
2024-06-26 — FDA grants accelerated approval to epcoritamab-bysp for relapsed or refractory follicular lymphoma

3 EMA oncology authorisations:

2024-06-17 — Truqap
2024-06-20 — Fruzaqla
2024-06-27 — Eribulin Baxter

8. Trade press¶

49 on-topic of 328 items across 3 feeds (n-stat 313, n-aacr-blog 11, n-acir 4). 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.

2024-06-28 — What Is a Cancer Vaccine? — n-aacr-blog
2024-06-28 — STAT+: FDA rejects Rocket Pharmaceuticals’ gene therapy for ultra-rare condition, another setback for field — n-stat
2024-06-27 — STAT+: House panel passes bills to expand Medicare obesity drug coverage, cancer screening — n-stat
2024-06-27 — STAT+: Are expensive whole-body MRI scans medically useful? A new clinical trial aims to find out — n-stat
2024-06-26 — STAT+: Journal retracts study tied to Micronoma’s effort to create cancer blood test — n-stat
2024-06-26 — STAT+: FDA issues long-awaited draft guidance for enrolling more people of color in clinical trials — n-stat
2024-06-26 — STAT+: Intellia says its CRISPR-based therapy can be redosed — n-stat
2024-06-26 — Three’s company: immune triads license CD8+ T cell-mediated antitumor immunity — n-acir
2024-06-25 — Cancer Today’s Summer 2024 Issue: Cardiovascular Toxicity, Immunotherapy in NSCLC, TILs, and More — n-aacr-blog
2024-06-25 — STAT+: Pharmalittle: We’re reading about Novo’s Wegovy approval in China, a Merck KGaA cancer trial failure, and more — n-stat
2024-06-25 — STAT+: Former Moderna executive launches genome editing company — n-stat
2024-06-24 — STAT+: Grail, aiming to market a blood test for cancer, faces host of challenges as it debuts on Nasdaq — n-stat
… and 37 more on-topic items.