ALXO — Q2 2026 Financial Report Analysis

ALX Oncology Q2 2026: a smaller loss, a much bigger bank balance, and a share count that more than doubled

ALX Oncology is a clinical-stage cancer drug developer: it has no approved products and has never had revenue from product sales, licenses or collaborations, so the quarter is about how much cash it spent, how much it has left, and whether its two drug programs are on schedule. In the quarter ended June 30, 2026, the net loss fell to $18.0 million from $25.9 million a year earlier, mainly because the company is running fewer clinical trials after it cut back its programs in 2025. Cash, cash equivalents and investments stood at $153.4 million, up from $48.3 million at the end of 2025, after a $150 million share sale in February 2026.

At a glance

• $153.4 million in cash and investments — enough, management says, to fund operations through the first half of 2028, which covers the mid-2027 data readout its main trial is built around.
• $34.7 million of operating cash used in the first half (about $17 million a quarter), down from $48.1 million a year earlier — the leaner, two-program company is spending roughly 28% less.
• Loss per share fell 73% to $0.13, but the dollar loss fell only 31% — most of the per-share improvement comes from having about 2.5 times as many shares, not from the business shrinking its losses that much.

The quarter in numbers

Figures from the company's Form 10-Q for the quarter ended June 30, 2026. "Non-GAAP net loss" is the company's own adjusted figure, which leaves out stock-based pay and several one-off items (detailed below).

Where the money went

Research and development (R&D) is almost all of what a company like this spends. It fell $4.9 million year over year to $13.1 million. The 10-Q breaks that down:

The company attributes the drop in clinical costs to a "change in clinical development strategy reducing the number of active clinical trials." In plain terms: in 2025 ALX cut staff (a workforce reduction in early 2025) and in August 2025 narrowed evorpacept development to one company-run trial in breast cancer, pausing a planned colorectal cancer study. Lower spending on older ("legacy") trials more than offset the cost of the new breast cancer trial and the first-in-human study of its second drug, ALX2004. Stock-based pay inside R&D rose $0.4 million only because the prior-year quarter was reduced by reversals tied to employees who left in the 2025 layoffs.

Over the first six months, R&D was $26.7 million versus $41.9 million, a 36% decline, with personnel costs down $4.5 million on top of the lower trial spending.

What the headline numbers hide

The per-share improvement is mostly dilution. On February 2, 2026, ALX sold 76,979,112 shares at $1.57 each plus pre-funded warrants — rights to buy shares for $0.001 each, essentially shares held in a different wrapper — for 18,574,120 more, raising about $140.4 million after $9.6 million of fees and expenses. Common shares outstanding went from 54.4 million at December 31, 2025 to 138.2 million at June 30, 2026, with another 13.7 million shares' worth of pre-funded warrants still outstanding (12.5 million from 2026, 1.25 million from 2023). That is roughly 152 million shares on a fully counted basis, about 2.8 times the 54.4 million common shares at year-end. The dollar loss narrowed 31%; the loss per share narrowed 73% because it is divided across far more shares. Anyone who held the stock before February owns a much smaller slice of the company.

The runway guidance is more conservative than current spending implies. At the first half's pace of $34.7 million of operating cash per six months (about $69 million a year), $153.4 million would last roughly 26 months from June 30, 2026, or into the second half of 2028. Management only guides to the first half of 2028, which suggests it expects spending to rise as the breast cancer trial completes enrollment and ALX2004 moves past early dosing. The 10-Q adds that this cash "may not be sufficient to fund all of the actions that are necessary to complete the development of evorpacept and ALX2004," so another raise before any drug reaches market should be expected. The company can also sell more stock through an existing at-the-market program (selling shares gradually on the open market).

Several small one-offs this quarter, a larger one last year. The year-ago loss included a $3.2 million impairment charge — a write-down of the value of leased lab space left idle after the 2025 preclinical layoffs. This quarter the company terminated its Palo Alto lease, recording a $0.2 million net gain on that line plus a $0.2 million loss on disposing of lab equipment and leasehold improvements. It also booked a $0.9 million loss on paying off its old loan early. Stripping out the year-ago impairment, the like-for-like decline in operating expenses is $5.2 million (to $18.3 million from $23.5 million excluding the lease line in both years), about 22% — still a real cut, just smaller than the headline 32%.

The adjusted (non-GAAP) loss excludes recurring stock pay. The company's non-GAAP net loss of $14.3 million adds back $2.7 million of stock-based compensation, which is a real, ongoing cost to shareholders through dilution, plus the one-offs above. The GAAP loss of $18.0 million is the better guide to the cost of running the company.

Interest income is a small but growing offset. With more cash invested, interest income rose to $1.5 million from $1.1 million. Most of the money now sits in short-term ($115.5 million) and long-term ($15.1 million) investments, with $22.7 million in cash.

Cheaper, more flexible debt. In June, ALX replaced its $10 million Oxford Finance/Silicon Valley Bank loan with a $10 million term loan from HSBC Ventures USA, with interest at the greater of the prime rate or 6.0%. It can draw another $20 million at its option through June 30, 2028, and up to $20 million more is possible ($10 million tied to development milestones, $10 million at HSBC's discretion). Debt is small relative to the cash pile: $9.7 million on the balance sheet against $153.4 million of cash and investments.

Takeaway: ALX traded ownership for time: the February share sale nearly tripled the share count but bought a runway that management says lasts into the first half of 2028, past the mid-2027 data from its 80-patient ASPEN-09-Breast trial. That readout will largely decide whether evorpacept has a future, and the cash means the company does not need to raise money before it.

The pipeline: what the company is betting on

Evorpacept blocks CD47, a protein many cancer cells display as a "don't eat me" signal to the immune system. It is designed to be given alongside existing antibody drugs, so those drugs' attack on the tumor also brings in immune cells. The company's case now rests on picking patients whose tumors have high CD47 levels:

• In its earlier ASPEN-06 gastric cancer trial, an exploratory analysis presented in November 2025 found that among patients whose tumors were still HER2-positive and had high CD47 (43 patients), adding evorpacept produced a 65.0% response rate versus 26.1% without it, with median progression-free survival (time before the cancer grew) of 18.4 months versus 7.0 months. In low-CD47 patients the gap was much smaller (37.5% versus 26.1%).
• In May 2026, at ESMO Breast Cancer, it presented an exploratory analysis of an older trial run by Jazz Pharmaceuticals combining evorpacept with Jazz's zanidatamab in heavily pre-treated breast cancer. All 5 patients with confirmed HER2-positive, high-CD47 tumors responded, versus 1 of 4 with low CD47. These are very small, after-the-fact subgroups, which can overstate an effect; the purpose of the new trial is to test the idea prospectively.
• ASPEN-09-Breast, the company's own Phase 2 trial of evorpacept with trastuzumab and chemotherapy in HER2-positive metastatic breast cancer after prior ENHERTU, dosed its first patient in January 2026. It was changed in August 2025 to a single-arm design (no comparison group) evaluated by CD47 level. Topline data from 80 patients are expected in mid-2027, and enrollment is described as on track.
• Partner-run studies continue: a Sanofi-sponsored trial in multiple myeloma is in dose optimization, and an MD Anderson investigator-run lymphoma trial reported a 92% complete response rate in untreated indolent non-Hodgkin lymphoma in December 2025.

ALX2004 is an antibody-drug conjugate (an antibody that carries a chemotherapy payload directly to cancer cells) aimed at EGFR, a protein found on many solid tumors. It entered a Phase 1 dose-escalation trial in August 2025; by January 2026 it had cleared its first two dose levels with no dose-limiting side effects and moved to a third cohort at 4 mg/kg. Initial safety data are expected in the second half of 2026. Earlier attempts at EGFR-targeted antibody-drug conjugates by other companies ran into side effects in healthy tissue, so safety is the first thing to watch.

Leadership changes

In June, board member Scott Garland became chairman, succeeding co-founder Corey Goodman, who left the board, and Michael Listgarten joined as general counsel.

Outlook

Management's stated milestones are ALX2004 initial safety data in the second half of 2026 and ASPEN-09-Breast topline data from 80 patients in mid-2027, with cash expected to last through the first half of 2028.

Our read: the next 12 months are mostly a waiting period financially. Quarterly losses should stay around $18 million, possibly rising as the breast cancer trial fills and ALX2004 adds patients; the gap between the current burn rate (implying runway into the second half of 2028) and management's first-half-2028 guidance suggests the company itself expects that increase. The near-term event is ALX2004's first safety data — clean results would give the company a second program to lean on if evorpacept disappoints. The decisive event is mid-2027: a single-arm trial cannot prove evorpacept beats standard treatment by itself, so the question is whether response rates in high-CD47 patients are strong enough to justify a larger, randomized trial, which would almost certainly require new funding or a partner. Things to check in the Q3 10-Q: quarterly operating cash use (to test the runway), any draw on the extra $20 million HSBC tranche, use of the at-the-market share program, and ASPEN-09-Breast enrollment progress.