Why Drugmakers Pay GeneDx to Give Away Genome Tests

Beren Therapeutics and GeneDx just launched a program called "Test for NPC" — genome sequencing offered at no charge to find people with Niemann-Pick disease type C, an inherited disorder that lets cholesterol build up inside cells and progressively destroys the brain. On its face this reads as philanthropy. It isn't. Someone is paying for every one of those tests, and following the money explains a lot about what GeneDx is actually worth.

Start with why a drug company foots the bill. NPC is caused by mutations in the NPC1 or NPC2 genes. It is, by any measure, tiny: prevalence is commonly put around one in 90,000 to 160,000 people. But it is likewise notoriously underdiagnosed, because its symptoms — clumsiness, trouble swallowing, developmental delay, psychiatric problems that first appear in adulthood — look like a dozen more common conditions. Studies put the average wait from first symptoms to a correct diagnosis at four years or more, and for the youngest patients the delay is longest. A diagnosis found early is an NPC patient who lives longer and on treatment.

That is the entire commercial logic. Beren's experimental drug adrabetadex is sitting with the FDA under a priority review now scheduled to decide by November 17, 2026; if approved it would be the first therapy to target the disease's underlying defect rather than just managing symptoms. A drug is only worth what its addressable market is worth, and for a company whose value is concentrated in a single rare-disease medicine, every undiagnosed NPC patient is a lost customer. Funding a free test to hunt them down is the most rational marketing expense in the industry.

The structure is what matters for GeneDx. In these sponsored programs the drugmaker pays the testing lab directly — no insurance claim, no out-of-pocket cost to the patient. In return the drugmaker gets de-identified patient data to study the disease, and GeneDx folds the same de-identified results into its own proprietary rare-disease database. Critically, this is the second NPC drugmaker to strike precisely this deal with GeneDx: Zevra Therapeutics, whose arimoclomol is already approved for NPC, launched its own sponsored exome sequencing program with GeneDx in March. Two competing drug companies, one testing lab, one diagnostic bottleneck. That is the tell.

So how much does one ultra-rare-disease program move GeneDx's revenue? Essentially none. NPC is too small for that, and no honest read of this announcement should pretend a handful of genome tests matters to a company booking $114.4 million of revenue in a single quarter. The value here is not the test volume; it is what the pattern reveals about the business.

GeneDx is in the middle of a transition best described as value migrating from the test to the data. The tests themselves are the reliable layer: exome-and-genome volume grew 32% year over year in the second quarter, gross margin was 70% on an adjusted basis, and the company just crossed into adjusted profitability, a rare state for a diagnostics lab. The flywheel is powered mainly by payer expansion — new state Medicaid coverage for sequencing and a commercial coverage deal opening access to tens of millions of lives. The sponsored-testing programs sit on top of that, doing double duty: booking paid volume without payer-collection friction while feeding the rarest, least reachable genomes into the database for free. That database is the "software" on top of the "hardware" — the asset that could someday set the multiple instead of the margin.

Here is where the question gets harder. The market has already heard this data story, run it up, and pulled it back. The stock sits near $86 after a move from a $32 low to a $170.87 high over the past year, still roughly a third below its high for 2026. Even at current levels GeneDx trades at more than five times trailing sales and a forward earnings multiple near 300 — not because it is richly profitable, but because it isn't yet. GAAP net loss in the second quarter was $17.7 million; the adjusted profit of $0.4 million is a rounding error dressed as a milestone.

That concentration of hope in a still-thin profit line is the real risk. Sponsored-testing revenue is lumpy and depends on drugmakers' continued willingness to pay, not on a durable payer contract. And the data layer — the reason to pay five times sales — is long-dated optionality, not today's revenue line; GeneDx still mostly makes money by running tests, not by selling its database. The free NPC test is a photograph of the thesis, not a proof of it.

The distinction a holder, or a watcher, should keep is not whether GeneDx can grow testing volume — the evidence says it can. It is whether the database becomes a monetizable asset or stays a cost of doing business. One sponsored test for an ultra-rare disease cannot answer that. But it shows you exactly which direction management is spending to find out.