AURA — Q2 2026 Financial Report Analysis

Aura Biosciences has no product sales. It is spending to reach one event: topline results from its Phase 3 CoMpass trial of bel-sar in early choroidal melanoma, a rare cancer inside the eye. In Q2 2026 the net loss grew 69% to $45.6 million. Most of that jump came from a one-time accounting cost tied to the company's change of CEO, not from higher day-to-day spending. More important, a $280.8 million share sale in May raised cash and securities to $323.8 million. After the quarter ended, Aura cut about 20% of its staff, scaled back its bladder-cancer program, and extended its cash runway guidance to "into the first half of 2029." Enrollment in CoMpass finished at 108 patients, more than the target, and topline data are still guided for the second half of 2027.

At a glance

• $323.8 million in cash and marketable securities (up from $144.2 million at the end of 2025). Management says this lasts into 1H 2029, which is about 1–1.5 years past the expected Phase 3 readout.
• $12.4 million of stock-based compensation in Q2, up from $3.8 million a year earlier. This non-cash cost, mostly from changing the terms of executives' equity awards during the leadership transition, explains about half of the $18.6 million increase in the net loss.
• Loss per share of $0.48, against $0.47 a year ago. The loss nearly doubled per quarter, but the average share count rose 64%, so each share carries about the same loss. That is the cost of funding through new shares (dilution).

The quarter in numbers

Aura is pre-revenue: it reported no revenue in either year, so revenue growth and operating margin don't apply. Percentage changes for the loss lines below show how much the loss widened.

Where the money went

R&D rose $7.9 million (+34%). The filing says this was "primarily due to ongoing clinical and CRO costs associated with the progression of the CoMpass global Phase 3 trial of bel-sar in early choroidal melanoma and manufacturing and development costs for bel-sar." A CRO, or contract research organization, is the outside firm that runs trial sites and data collection. This is the spending that matters: it pays for the trial whose result decides whether Aura has a product. Stock-based compensation within R&D was flat ($2.1 million in both quarters), so this increase is real operating spend.

G&A tripled, from $5.7 million to $17.3 million. Most of the increase is non-cash. Stock-based compensation charged to G&A went from $1.8 million to $10.3 million. The filing says the rise was "primarily driven by increased stock-based compensation expense resulting from equity award modifications in connection with executive leadership transitions, as well as higher professional fees." The transition is the change at the top: founder Elisabet de los Pinos stepped down as CEO on April 30, 2026, and Natalie Holles took over. The 10-Q lists a separation agreement and an amended consulting agreement with Dr. de los Pinos, plus a new inducement equity grant to Ms. Holles: an option on 2.17 million shares, 0.60 million restricted stock units and 0.55 million performance-based units. Without stock-based compensation, G&A was about $7.0 million against $4.0 million a year ago. That is still higher, and the filing points to professional fees.

What the headline numbers hide

• The loss overstates cash spending. Six-month net loss was $79.3 million, but cash used in operations was $62.4 million. The $16.9 million gap is almost all stock-based compensation ($16.2 million), which costs shareholders through dilution rather than costing the company cash. Without stock compensation, the Q2 net loss was about $33.3 million against $23.2 million a year ago: still 43% wider, but much less than the 69% headline.
• Implied Q2 cash burn was about $33 million. Cash and securities were $114.7 million on March 31 (per the Q1 release). Add the $280.8 million offering and subtract the $39.0 million repurchase described below, and Q2 burn works out to roughly $32.7 million, in line with the first-half average of about $31 million per quarter.
• Part of the May raise went straight back out. Aura sold 46.1 million shares at $6.00, plus pre-funded warrants for 3.8 million more, for $299.4 million gross and $280.8 million after fees. A pre-funded warrant is a near-fully-paid right to receive shares, used by investors who don't want to hold too many shares directly. Two days later it spent about $39.0 million buying back 6.92 million shares at $5.64 from Matrix Capital Management, an existing holder; the Q1 release says this was all of Matrix's stake. Net, the company added about 43 million shares (counting the pre-funded warrants) for $241.8 million of cash it can use.
• Dilution keeps rising. Shares outstanding went from 63.6 million at year-end to 103.5 million on June 30. There are also 7.37 million pre-funded warrants and 3.80 million common warrants exercisable at $4.90. The flat loss per share reflects the bigger share count, not a smaller loss.
• More one-time costs are coming in Q3. The August restructuring will cost an estimated $2.9–3.2 million, covering severance and accelerated vesting of some stock awards, with "a significant majority" paid in Q3 2026. The CFO (Tony Gibney), Chief Legal Officer and CTO are leaving. Mr. Gibney receives nine months of salary continuation, and his equity keeps vesting through a consulting agreement that runs to May 2027. Q3 G&A will therefore carry another round of transition costs.
• The cash statement is not a going-concern warning. The 10-Q says cash will last "at least 12 months" from the filing date. That is the standard wording, not a doubt about survival. Management's own forecast is into 1H 2029.

Did the last guidance hold up?

This is our first report on Aura, so there's no earlier report to check. Against what management said in the Q1 2026 release (May 11), every operating milestone was met or beaten. CoMpass enrollment was "expected by mid-2026" and finished at 108 patients, more than the target. Initial 3-month bladder cancer (NMIBC) data were expected "mid-2026" and arrived in August. Runway guidance moved from "into the second half of 2028" (which already counted the May offering) to "into the first half of 2029." The one change in direction is NMIBC. Its data looked encouraging: an 81% objective response rate and a 69% complete response rate at 3 months among 16 intermediate-risk patients, with only Grade 1 (mildest) treatment-related side effects. Even so, management is cutting its funding. Aura says it will finish the protocol's 12-month follow-up "to preserve optionality for value creation in the context of future potential strategic discussions," which most plausibly means a partner or licensee.

Takeaway: Aura now has enough cash to reach its only near-term make-or-break event, the CoMpass topline readout in 2H 2027, without raising money first. At the first-half spending pace (about $31 million a quarter), it would still hold roughly $130–140 million when the data arrive, by our estimate. The price was diluting existing holders by about two-thirds more shares and stepping back from its bladder-cancer program. From here, the investment case depends almost entirely on one trial.

What it means for patients and the market

Early choroidal melanoma is currently treated with radiation: plaque brachytherapy (a radioactive disc sewn onto the eye) or proton beam therapy. The filing notes both often lead to "significant vision loss, and potential legal blindness in the treated eye." The alternative is removing the eye. Bel-sar is a light-activated drug injected into the eye that binds to tumor cells. Its pitch is tumor control while keeping vision. In Aura's earlier Phase 2 study, 8 of 10 Phase 3-eligible patients had tumor control, and 9 of 10 kept their visual acuity. That is a small sample, which is why the 108-patient CoMpass trial is the real test. CoMpass compares bel-sar with a sham (placebo) procedure under a Special Protocol Assessment, an advance agreement with the FDA on trial design and analysis, which reduces, though it doesn't remove, the risk of a dispute over the results later.

Outlook

Management's milestones:

• CoMpass topline data (15-month primary endpoint): 2H 2027, unchanged.
• Updates on bel-sar in metastases to the choroid and in cancers of the ocular surface, including guidance on when those studies will finish: Q1 2027. Aura says it is increasing resources for both.
• Restructuring substantially complete by the end of Q3 2026.
• Cash runway: into 1H 2029.

Our read: The runway math holds up. $323.8 million at about $31 million a quarter is roughly ten quarters, which reaches late 2028 even before any savings from the 20% staff cut, so "into 1H 2029" assumes only modest savings. Two caveats. First, the 10-Q itself says the cash "may not be sufficient to fund bel-sar through regulatory approvals." Preparing a launch after positive data would almost certainly need another raise or a partner. Second, the next quarter will look noisy: expect Q3 G&A to stay high from restructuring charges and executive exits, then fall in Q4 as the smaller team takes effect. In the 12 months before the 2H 2027 readout, the useful signals are steady R&D spending (a sign the trial is running normally), quarterly burn staying near or below $30 million after restructuring, and the Q1 2027 update on the two smaller eye-cancer programs.