CNSY — Q2 2026 Financial Report Analysis

Spending tripled while the test business is still at the starting line

Cerenome, Inc. (Nasdaq: CNSY) is the company formerly called Plus Therapeutics. It changed its name and ticker on August 3, 2026, after the quarter ended, to reflect a shift from a pure drug developer to a company built around brain and spinal-cord ("CNS") cancers through three businesses: CNSide, a lab test that counts tumor cells in the fluid around the brain and spine; REYOBIQ (rhenium-186 obisbemeda), an experimental radioactive drug injected directly into brain tumors; and a data/AI effort that is still early. It is the same legal entity with the same SEC filing history, not a merger.

In the second quarter of 2026 (April–June), Cerenome lost $9.0 million, against a $5.2 million profit a year earlier. That swing is mostly an accounting mirage on the 2025 side (explained below). The real change is spending: operating expenses rose from $2.9 million to $9.5 million as the company hired a commercial team for CNSide, paid more lawyers and advisers, and ran its clinical trials. Meanwhile its only reported revenue, grant money from a Texas cancer-research agency, fell 70%. CNSide began billing insurers this year, but the company judged the amount collectable so far too small to record as revenue.

At a glance

• $9.5 million in quarterly operating expenses, up 223% from $2.9 million. Research spending rose 242% and overhead 209%. This is now a company spending at roughly 3x last year's pace, on purpose.
• $8.6 million of cash and short-term investments on June 30, against $13.4 million of cash burned by operations in the first half alone. At that pace the June balance covers about four months, which is why the filing carries a formal "going concern" warning (auditor-speak for "it is not certain the company can keep operating for another year without raising money").
• 232 CNSide tests in the first half of 2026, about 39 a month. Management's 2026 goal is an annualized run-rate above 1,250 orders, roughly 104 a month, so test volume needs to more than double from its first-half pace.

The numbers

All figures in millions of US dollars except per-share amounts. Source: Cerenome's Form 10-Q for the quarter ended June 30, 2026.

Share counts reflect the 1-for-25 reverse stock split (every 25 old shares became 1 new share) that took effect April 2, 2026. "Operating margin" is not a useful measure here: costs are more than 20 times revenue.

For the first six months, grant revenue was $1.44 million (down 41% from $2.45 million), the operating loss was $16.2 million (vs $5.1 million), and the net loss was $16.0 million, or $2.36 per share (vs $12.3 million, or $12.54 per share).

Where the money went

Research and development rose $3.0 million year on year to $4.3 million. The filing breaks the increase down as $1.0 million more for clinical trials, $0.7 million more in R&D staff pay, $0.7 million more for outside research services, $0.3 million for the diagnostics side and $0.3 million of other costs. Quarter on quarter the jump is also sharp: R&D was about $2.9 million in Q1 2026, so Q2 was roughly 49% higher.

General and administrative costs (overhead: executives, legal, rent, finance) rose $3.5 million to $5.2 million. The biggest pieces were $1.6 million more in legal and professional fees and $1.2 million more in pay, of which $0.6 million was stock-based compensation (paying staff in shares or options rather than cash). Rent added $0.4 million. G&A was roughly flat versus Q1 2026 ($5.3 million), so the overhead step-up happened at the start of the year and has held.

Grant revenue fell because it is a reimbursement, not sales. The Cancer Prevention and Research Institute of Texas (CPRIT) pays about two-thirds of the cost of developing REYOBIQ for leptomeningeal metastases (cancer that has spread to the lining of the brain and spinal cord), up to $17.6 million. Cerenome had received about $15.9 million of that by June 30, so this source is close to used up. Less spending on that specific trial in the quarter meant less reimbursement.

What the headline numbers hide

• Last year's "profit" was not real operating income. Q2 2025's $5.2 million net income came from a $6.5 million non-cash gain on the changing value of warrants (rights investors hold to buy shares at a set price) issued in a March 2025 financing. Strip that out and Q2 2025 was a loss too, which is why that quarter's diluted EPS was -$0.16 even though basic EPS was +$0.62. Comparing a $9.0 million loss to a $5.2 million profit overstates how much things deteriorated; comparing operating losses ($9.1 million vs $1.5 million) is the honest view. Q2 2026 itself had no such items: other income was just $51,000, mostly interest.
• Cash burn was flattered by paying suppliers later. First-half operating cash outflow ($13.4 million) was smaller than the $16.0 million net loss. Part of the gap is legitimate non-cash cost ($1.9 million of stock compensation), but $1.8 million came from accounts payable and accrued expenses rising, from $5.9 million to $7.6 million. Bills that were owed but not yet paid will need cash later. Working the other way, $1.4 million of grant money had been earned but not yet received (the grant receivable rose from $0.3 million to $1.8 million).
• "Flat cash" isn't flat. Cash plus investments was $8.6 million at both year-end and June 30, but only because the company raised money in between: about $13.9 million net from a January 2026 share-and-warrant sale and $1.8 million net from selling shares into the market (an "at-the-market" program). It also paid $4.5 million back to March 2025 investors to cancel 272,821 shares under an earlier agreement; that money had been set aside as restricted cash at year-end.
• Per-share figures are distorted by dilution and the reverse split. Weighted basic shares rose from 1.9 million to 6.9 million year on year (split-adjusted). Shares outstanding went from 5.5 million at year-end to 7.3 million at June 30, a 32% increase in six months. Any per-share comparison has to be read with that in mind.
• Diagnostic revenue exists but isn't booked yet. The company started billing for CNSide tests in the first half but recorded none of it, because under its accounting policy it only books revenue it is confident it will collect, and with little payment history from insurers the estimated amount was "immaterial." That is conservative, but it also means there is no revenue figure yet to judge the test business by.
• No guidance change to check. Management affirmed the milestone list it set out with its full-year 2025 results; it gives no revenue or expense guidance.

Takeaway: Cerenome has turned itself into a commercial diagnostics company in its spending before it has become one in its revenue: operating costs ran at $9.5 million a quarter while CNSide revenue was too small to record, and the $8.6 million cash balance at June 30 was well short of what six months of operations consumed ($13.4 million). Everything now depends on CNSide test volumes and insurer payments ramping fast enough to reduce reliance on repeated, dilutive fundraising.

What changed after the quarter

• New secured, convertible debt (September 2026). On September 4 the company agreed to sell up to $21.3 million (face value) of senior secured convertible notes to one institutional investor, with an initial closing expected to bring in $3.0 million gross. The terms are expensive: a 6% original issue discount (the company receives less than the face amount it owes), 8% annual interest guaranteed to maturity, a one-year term, conversion into shares at $2.74 (with a $0.50 floor if the price is reset), security over substantially all assets of Cerenome and CNSide, and a royalty of 2.5% of CNSide's gross revenue each quarter while the notes are outstanding. The larger tranches depend on conditions, including registering the conversion shares. Terms like these usually signal limited access to cheaper funding.
• Bigger Houston lab space, later. An August lease amendment more than tripled the planned Houston space (from about 11,400) to about 36,500 square feet, with base rent rising to about $188,000 a month, but pushed the lease start to around November 2027 with the first rent payment due around June 2028.
• Stock exchange listing. Nasdaq is monitoring the company's minimum stockholders' equity requirement through August 22, 2026. Equity was $5.5 million at June 30, and a single quarter's loss was $9.0 million, so staying compliant depends on continued capital raising. The company regained compliance with the minimum share price rule in April after the reverse split.

Outlook

Management's milestones for the rest of 2026: define the best dose and dosing interval for REYOBIQ in leptomeningeal metastases (about a third of patients in the multi-dose trial were enrolled by June 30, with no dose-limiting side effects so far); finish enrollment in the ReSPECT-GBM Phase 2 trial in recurrent glioblastoma, with data expected in Q1 2027; dose the first patient in the pediatric brain cancer trial (expected Q3 2026); win Medicare coverage for CNSide (the lab already has a Medicare billing number and its own billing code since July 1); expand contracted insurer coverage beyond 150 million people; and exceed an annualized run-rate of 1,250 test orders.

Our read: management also expects both R&D and G&A to keep rising for the rest of 2026. With operating costs already near $9.5 million a quarter, the next two quarters will likely consume more than the June cash balance plus the $3.0 million initial note proceeds, so further share sales or larger note draws are close to certain. Two things to watch in the Q3 report (10-Q due around mid-November): whether CNSide revenue starts appearing on the income statement, and the monthly test pace against the roughly 104 per month implied by the 1,250 run-rate goal. The glioblastoma Phase 2 data in early 2027 is the bigger single event for the drug side of the business.