CBIO — Q2 2026 Financial Report Analysis

Crescent Biopharma lost $24.8 million in the second quarter of 2026, compared with $21.8 million a year earlier. The extra loss came from research spending: R&D rose 60.5% to $19.4 million as its lead cancer drug, CR-001, moved into a global human trial and manufacturing work stepped up. Crescent has no product sales, so for now the useful questions are how fast it spends cash and how long the cash will last. After the quarter ended it raised a net $133.5 million in a share sale, taking pro forma cash to $305.1 million. Management now says that money lasts into the second half of 2028, past the first clinical data readouts expected in 2027.

At a glance

• $171.6 million in cash at June 30, $305.1 million pro forma (counting the July share sale). This cash is the company's lifeline until a drug is approved or partnered.
• Cash fell $17.6 million in the quarter (from $189.2 million at March 31). That is a quarterly burn of roughly $18–23 million, depending on how one-off items are counted.
• About 30% more shares. Shares plus share equivalents rose from 33.5 million to 43.5 million pro forma after the July offering, so each existing share now owns a smaller slice of the company.

The quarter in numbers

Source: Form 10-Q for the quarter ended June 30, 2026. Pro forma cash and share counts come from the company's July 30, 2026 earnings release.

What Crescent is developing

Crescent became a public company in June 2025 through a reverse merger with GlycoMimetics. In a reverse merger, a private company takes over a listed shell so it doesn't need a traditional IPO. Crescent has three cancer drug candidates:

• CR-001: a "PD-1 x VEGF bispecific antibody", one molecule built to hit two targets. It blocks PD-1, the brake on the immune system that Merck's Keytruda also blocks, and VEGF, a signal tumors use to grow blood vessels. CR-001 is designed to copy the binding properties of ivonescimab, a rival drug whose early clinical data Crescent points to as validation. The global Phase 1/2 trial (ASCEND) started in Q1 2026, and the Q2 release says enrollment "continues to progress". Partner Kelun-Biotech received Chinese approval to start trials (IND clearance) in May 2026. Kelun has since begun a Phase 2 trial in China combining CR-001 with its approved drug sac-TMT in about 206 lung cancer patients.
• CR-002: an antibody-drug conjugate (ADC), an antibody that carries a toxic payload straight to cancer cells. It targets PD-L1. The company planned to file its US trial application in mid-2026 and start a global Phase 1/2 trial in the second half of 2026.
• CR-003 (SKB105): an ADC aimed at ITGB6, licensed from Kelun in December 2025 for markets outside Greater China. A Phase 1/2 trial is running in China.

Why the loss widened

The whole increase in operating expense came from R&D. G&A was roughly flat.

The 10-Q says external R&D rose $4.3 million, to $12.5 million, "primarily related to an increase in CMO development and manufacturing expenses and an increase in our CRO expenses related to ongoing clinical trials and toxicology studies." A CMO is the outside firm that makes the drug and a CRO is the outside firm that runs the trials. Lower payments to Paragon, the related-party lab that discovered Crescent's antibodies, partly offset the increase after the ADC option agreement with Paragon ended. Paragon-related R&D was zero this quarter, against $5.0 million a year ago. Spending has moved from paying an affiliated discovery lab to paying for trials and drug supply. That is the normal shift as a biotech moves from lab work into human trials.

G&A slipped $0.2 million. Stock-based pay fell $1.4 million because of award forfeitures in the prior year, and consulting fees fell $0.4 million as staff were hired in-house. Higher office costs ($0.7 million) and other pay partly offset those savings.

Takeaway: The quarter's loss is not the number to watch. The cash math is: Crescent spends roughly $20–23 million a quarter today, but its new runway guidance only works if spending climbs to about $30–35 million a quarter. The $133.5 million raised in July pays for a larger trial program, not only for more time.

What the headline numbers hide

• A $1.8 million one-off gain cushioned the loss. "Other income" of $1.77 million came from selling legacy GlycoMimetics intellectual property. Without it, the Q2 net loss would have been about $26.6 million, 22% wider than a year ago instead of the headline 14%.
• Interest income helped too, but it shrinks as cash is spent. Interest on the cash pile was $1.5 million, against $0.3 million a year ago. The year-ago quarter also carried $1.1 million of interest on convertible notes that have since turned into shares. These below-the-line items made the net loss look about $2.3 million better than the operating trend alone.
• The six-month cash flow is flattered by a one-time collection. Operating cash outflow for the first half was $28.4 million, much smaller than the $48.1 million net loss. Most of the gap is an $18.0 million receivable that Kelun paid in Q1 under the CR-001 license, plus $8.1 million of non-cash stock compensation. Without the $18.0 million collection, first-half operating cash use was about $46.4 million, or roughly $23 million a quarter. Two other outflows sit outside "operating" cash: $8.0 million of withholding tax paid on Kelun's behalf for the SKB105 license, and $7.1 million of offering costs.
• Per-share loss is not comparable year over year. Loss per share improved from $(4.93) to $(0.74), but the business did not get better. The year-ago figure was spread over only 3.9 million weighted shares because the reverse merger closed in mid-June 2025. This quarter it was spread over 30.5 million. A smaller loss per share here mostly reflects more shares.
• Dilution is the cost of the runway. The July offering sold 9.39 million shares at $14.50, plus pre-funded warrants for 0.53 million more. (A pre-funded warrant is a near-free option to take a share later.) Gross proceeds were $143.7 million and net proceeds $133.5 million. The price was above the $13.41 paid in the December 2025 private placement. Crescent also set up a $200 million "at-the-market" program on July 1, which lets it sell shares into the market over time. The 10-Q says none had been sold yet, but it is a ready channel for more dilution.
• Guidance moved less than the new money suggests. In April, with $189.2 million, management said cash would fund operations "into 2028". Now, with $305.1 million pro forma, it says "into the second half of 2028". About $134 million more cash added only a few quarters to the stated runway. That points to a much higher planned spending rate. Management says R&D and G&A "will continue to increase significantly". By our math, funding July 2026 to late 2028 (about 27–30 months) with $305 million implies roughly $10–11 million a month, against about $7–8 million a month now.
• No revenue to speak of. Q2 license revenue was zero. The first half's $1.0 million (all in Q1) was a slice of Kelun's $20.0 million upfront payment for CR-001, recognized over time. $3.1 million of that payment is still on the balance sheet as deferred revenue.
• The Q1 10-Q amendment was administrative. The 10-Q/A filed June 22 only added a missing disclosure about executives' Rule 10b5-1 trading plans. No figures were restated.

What to watch next

The financial trajectory is set. Losses will grow from here as ASCEND enrolls more patients and CR-002 enters the clinic. Whether that spending pays off depends on 2027 data. The company's stated timeline:

Our read: the balance sheet is no longer the near-term risk, because the July raise covers every readout listed above. The two things to track in the next 10-Q (Q3 2026, due around early November) are both checkable. First, did CR-002 actually start its trial in the second half as promised? Second, does quarterly R&D move toward the roughly $25 million-plus level the runway guidance implies? A burn that rises more slowly would mean either a longer runway or delays in trial timelines. The filing alone can't tell those apart, so the trial-start updates matter as much as the dollar figures. The real test is the Q1 2027 ASCEND data. CR-001's pitch is that it reproduces ivonescimab's pharmacology, and that claim can only be confirmed in patients.