Leal Therapeutics Patent Portfolio Analysis

Thirty million, and a twenty-month silence

A $30M Series A, with Eli Lilly, OrbiMed and Dementia Discovery Fund on the line. For a company founded in 2021 that is a serious set of names, and the kind of syndicate that reads the science before it reads the deck. The money is the visible event. It is also the least informative thing here, because money says what three investors concluded, not what the company built to get them there.

The more interesting date is the one before it. The last public filing from Leal Therapeutics is 17 December 2024, twenty months ahead of this round. Anything filed since then is still inside the publication delay and invisible. So the record you can read stops well short of the round. Everything below is the portfolio as it stood when the diligence started, not as it stands now.

Seven pending, nothing yet enforceable

Seven patent families. Zero granted. Seven published and pending, filed through EP and WO routes. The granted ribbon is not thin, it is absent. First grants are expected around 2027, which means the entire portfolio is currently a set of claims under examination rather than a set of rights.

That is normal for a company this young and it is also the honest reading: there is nothing to lean on yet. What the pending stack does carry is dates. Priority sits in 2024 across all seven, which fixes the position in time whatever the examiners eventually allow. The count is settled. The scope is not, and scope is the only part of this that will matter in 2027.

The granted ribbon is not thin, it is absent.

Small molecules, filed in biology's classification

Two bands. The claims cluster in enzymes, cells and genetic engineering (C12N) and in medicinal preparations (A61K). Set that against the filing titles: compositions and methods for modulating SPTLC2, compositions and methods for modulating APOE, small molecule inhibitors of glutaminase. Two named genetic targets and one enzyme inhibitor, which is exactly why the record lands where it does.

The pairing is worth holding onto. A61K is where the therapeutic composition claims live. C12N is where the biology of the target lives. Filing across both, in the same year, on the same three subject areas, describes a company writing around targets rather than around a single chemical series. Sitryx Therapeutics has cited this work twice already, which is early but is a real external read on the same ground.

One name on every single family

Eight inventors are named across the seven families. Asa Abeliovich appears on all seven. Laura Heckman appears on all seven. Herve Rhinn appears on six. The remaining five names sit at the edges of the diagram, attached to one or two families each.

This is a founder-scientist portfolio, not a distributed research operation. Three people account for essentially the whole record, and two of them missed nothing. That concentration is a strength in the sense that the scientific thesis is coherent and traceable to specific individuals. It is also a dependency, and an honest one to name: the intellectual thread through SPTLC2, APOE and glutaminase runs through a very small number of heads.

Two people missed nothing across the entire portfolio.

Seven families in a very populated room

On the same scale, over the same five years, in the same CPC groups: University of California at 682 families, Zhejiang University at 476, China Agricultural University at 454, Jiangnan University at 369, Huazhong Agricultural University at 325, China Pharma University at 285. Leal Therapeutics has seven. The circles are not close.

That comparison is less damning than it looks, because most of that volume is university output across an enormous classification space rather than competing therapeutic programs. But it does settle one question. Count is not the asset here. In a space this populated, the only thing that separates seven families from noise is how narrowly and how specifically the allowed claims end up drawn around SPTLC2, APOE and glutaminase.

What the next publications have to show

Put the pieces together. A complete portfolio filed inside 2024, nothing since that is visible, three targets, three inventors doing nearly all the work, and a hiring plan of twenty roles that is sixteen parts biology and clinical. The hiring matches the filings exactly. That coherence is the finding: this is a company doing one thing, and the record says so without ambiguity.

So the moat is real in shape and unproven in strength. It is a dated, coherent, target-led position with no grants behind it and a crowded classification around it. The twenty-month publication gap is where the answer lives. Whatever was filed after December 2024 will either extend those three target families or open a fourth front, and that difference is the whole story.