Market Atlas · Cell therapy · early access
Every immune cell therapy trial, center, company, approval and 10,000 recent papers, on one map. Free in early access. Access is reviewed by hand; we email you a sign-in link once you’re approved.
We use your details to review your request and to tell you about updates. No spam; no selling.
What should we track for you?
Adoptive cell therapy · T cell · NK · TIL · CAR · ClinicalTrials.gov through 2026
Every institution running a T cell, NK, TIL or CAR trial, where it is, what it runs, who sponsors it, and what the NIH paid for nearby. Built from the trial registry, not from a vendor list.
Walking back from 2026
Trials started per year, one line per modality, with FDA approvals marked underneath. Hover for the numbers. Lines respond to the map filters below.
The map
One dot per institution, sized by its trials and papers, coloured by the category it runs most. Drag to move, scroll or pinch to zoom, or use the buttons. Hover a dot for detail; click it for its trials and papers. Search finds trials and papers by keyword, and the map narrows to the institutions behind them. Use the side table to drill from country to state, city and institution. Filters narrow the map, the side table, the trend chart and the centers tab at once.
Readouts
Pick any mix. The readout is computed from the trials and papers above: how much activity, how far along, which targets, who is doing it, and whether the evidence is in vitro, in vivo or clinical. In vivo means the cells are engineered inside the patient rather than made in a lab.
Institutions
Site names were rolled up to institutions, so spelling variants of one hospital count once. NIH column is fiscal 2025 cell-therapy-matching awards to that organisation, US only by construction. Click a row to open its trials, papers and methods underneath; click again to close.
| Institution | City | Country | Trials | Papers | CAR | Runs | NIH FY25 |
|---|
Directory
One row per company, from six sources: trials it sponsors, SITC and ASH abstracts, AACR and ASCO abstracts filtered to cell therapy. Click a row for its website and a link to every trial, poster and abstract. Built quick: companies are detected from affiliation text, so expect some misses and the odd department that reads like a company. ASCO lists no affiliations, so a company is credited there only through a trial it sponsors or its name in the abstract. Trials include 124 the registry filter missed, because they were filed under the wrong treatment type (Arcellx anito-cel as a biologic, Adicet ADI-001 as a small molecule). They are found here by product name or a lymphodepletion regimen, and show in company rows but not on the map. Vendors, CROs and device makers sit under Type: Tools & services.
| Company | Country | Stage · on market | Trials | SITC | ASH | AACR | ASCO | Mesa |
|---|
Early signal
Three layers, read together. SITC is immunotherapy science, so a company there is doing preclinical and translational work. ASH is where cell therapy reaches patients, since half these trials are haematological and multiple myeloma is the single biggest condition. The registry is what has been filed. A company at SITC with nothing at ASH and no registered trial is an early signal: work that is real but has not reached a patient or a filing yet.
Below: the early-signal companies. Two years rather than one is the point, because a company appearing in both is running a sustained programme rather than presenting once.
| Company | SITC abstracts | Years |
|---|
Development pipeline
Preclinical counts papers (2023-2026, in vitro and in vivo work); Phase 1 to 3 count trials; on market counts FDA-approved products. Each stage has its own unit, so compare modalities within a stage, not stages with each other. Trials and papers follow the map filters.
How it was built, and what it misses
Scope. registered trials were filtered to adoptive cell therapy by matching title and conditions against a category vocabulary, then unioned with the registry's own modality label. Matching runs over title, conditions and intervention names. That yields in-scope trials. A further cell-therapy trials were found and deliberately excluded: mesenchymal stromal cells, haematopoietic transplant, dendritic cell and islet work. They are cell therapy, but not T, NK or TIL.
Two recall bugs, both caught after the first build. The TIL long form was compiled case-sensitively, so "Tumor-infiltrating lymphocyte" never matched: TIL went from 89 to 200. Worse, the first pass searched only title and conditions, so brand-name products were invisible. A trial called "A Study of Lisocabtagene Maraleucel" says CAR nowhere. Adding intervention names moved Kite from 8 trials to 31 and Novartis from 7 to 23, and added 472 trials overall. Counts are a floor, not a ceiling.
Funding is the weak layer. Of companies, only 63 matched an entity carrying deal data, and a second pass against 3,399 SEC filers added none. The matches skew to large acquirers. Cabaletta, Precigen, Cartesian, NexImmune, Luminary and Miltenyi Biomedicine are all absent. Company funding on this page should be treated as missing, not as zero.
Institution counts are a floor. Site names are merged in two narrow steps: one rule per named US center, scoped to its city (MD Anderson in Houston, not Banner MD Anderson in Arizona), then names that are identical once punctuation and “(Adults only)” tags are removed. That took 3,606 names to 3,042 and MD Anderson from 104 trials to 172. Smaller centers outside those rules can still be split. A broad merge was tried earlier and reverted, because a shared place-word fused Seattle Children’s with Washington University. A further anonymous registry placeholders ("Local Institution 0023") were dropped rather than mapped. No trial summaries. The extract carries titles, conditions and interventions, but not ClinicalTrials.gov brief summaries. Every trial links out to its own record instead. NIH is one year and one country. Fiscal 2025 only, matched by organisation name, so it covers US academic centers and nothing else. It is a floor on US academic funding and says nothing about anywhere else.