AUTL — Q2 2026 Financial Report Analysis

Overview

Autolus Therapeutics, the London-based maker of the CAR T-cell therapy AUCATZYL (obe-cel) for adults with relapsed or refractory B-cell acute lymphoblastic leukemia (B-ALL, a fast-moving blood cancer), more than doubled its quarterly sales in Q2 2026. Total revenue reached $45.7 million, up 118% from $20.9 million a year earlier and up 74% from $26.2 million in Q1 2026. The bigger change was in cost: for the first time the company made a solid profit on each dose sold, with a product gross margin of about 55% (sales minus the direct cost of making the therapy, as a share of sales), compared with 6% in Q1 and a negative margin in every quarter of 2025. The company is still losing money overall, with a net loss of $39.1 million against $47.9 million a year ago, because sales, administration and research costs still exceed what each dose earns.

Autolus reports in US dollars under US GAAP and files standard 10-Q/10-K reports, even though it is a UK company.

At a glance

• $45.7 million revenue (+118% YoY): more patients were treated at existing US cancer centers, new centers came online, and the UK launch (January 2026) added its first full sales quarter.
• ~55% gross margin vs. negative a year ago: manufacturing cost per batch fell as volumes rose and April's cost cuts took hold. Only about $0.4 million of the benefit came from "free" pre-approval inventory, so the improvement is mostly real.
• $201.6 million cash and securities at June 30, plus $75 million borrowed in July. Management says this funds operations into Q2 2028, a forecast that assumes it also draws a further $25 million of debt and that sales keep growing.

Key metrics

n/m = not meaningful. Gross margin is calculated as net product revenue less cost of sales, divided by net product revenue, the same definition the company uses. Q2 2025: ($20.9M − $24.4M) / $20.9M.

What drove the quarter

Demand. The 10-Q attributes the $24.7 million sales increase "primarily" to "an increase in the number of AUCATZYL doses administered to patients." Revenue is recorded only once a patient has received the second and final infusion, so the figure tracks completed treatments, not orders. The company says growth came from deeper use at authorized treatment centers it already had, new centers coming online, and UK sales, which have run through the NHS and private centers since January 2026. The filing does not split US and UK revenue.

Manufacturing cost. This was the main story. Cost of sales fell to $20.5 million from $24.4 million even though sales more than doubled, so cost of sales dropped from 117% of revenue to 45%. Management credits lower cost per batch from higher volume and "operational changes at the Nucleus facility" (its commercial manufacturing site in the UK), the April 2026 efficiency program, and lower inventory write-downs ($1.2 million vs. $2.5 million). Third-party royalties within cost of sales rose to $1.6 million from $0.6 million as volumes grew. Autolus says its long-run margin target for adult ALL is 65–70%.

Operating costs. R&D was flat at $27.9 million. A $3.6 million drop in R&D staff costs was offset by $1.5 million more spending on clinical trials and clinical supply, plus $2.6 million less in UK R&D tax credits, which are netted against R&D and shrank after the company moved to a less generous UK credit scheme. SG&A rose $10.9 million to $41.2 million. Of that increase, $7.4 million was employee costs, "mainly driven by" severance from the April layoffs and more commercial staff. The rest was IT and facilities ($2.1 million) and professional fees ($1.1 million).

What the headline numbers hide

• About $2.7 million of the net loss improvement is an accounting re-estimate. Autolus owes future royalties and milestones to Blackstone and BioNTech from past funding deals (a $284.3 million liability on the balance sheet). The company reports interest on that liability, and this quarter changed assumptions in its valuation made the line positive $2.7 million (Q2 2025: positive $6.8 million). Without it, the Q2 pre-tax loss would have been about $41.3 million rather than $38.6 million. The same line was an $8.5 million expense for the first half as a whole, so it can swing in either direction from quarter to quarter.
• The year-over-year net loss comparison understates the operating improvement. The operating loss shrank by $17.4 million, but the net loss shrank only $8.8 million. Interest income on cash fell to $1.9 million from $5.2 million because balances are smaller, and the Blackstone/BioNTech re-estimate credit was smaller than last year's.
• One-off costs inside SG&A. Severance and termination costs from the 13% workforce cut totaled $4.2 million in the first half of 2026 (about $8.0 million expected over the whole program, including $2.4 million booked in 2025). Management expects the cuts to save about $15 million a year from 2027. That is meaningful, but small next to an H1 operating loss of $103.4 million.
• Cash burn vs. reported loss. Operating cash outflow for the first half was $90.4 million, against a net loss of $110.7 million and $148.3 million of cash outflow in H1 2025. Cash and securities fell $27.8 million in Q2, from $229.4 million to $201.6 million, compared with roughly $71 million in Q1 (from $300.7 million at year-end). The company says Q2 cash use was lowered by a UK R&D tax credit receipt, so Q2's burn is not a clean run rate.
• Receivables are growing with sales, not ahead of them. Accounts receivable rose to $44.4 million from $24.0 million at year-end, which is roughly one quarter of sales outstanding. That is consistent with the jump in revenue, not a warning sign by itself.
• Shareholders' equity is thin. Book equity fell to $74.8 million from $178.1 million at December 31 as losses accumulated. Liabilities total $418.4 million, mostly the royalty obligation and leases, and that excludes the new $75 million secured loan drawn in July, which will appear from Q3.

Guidance and funding

Autolus raised its full-year 2026 AUCATZYL net product revenue guidance to $140–150 million, up from $120–135 million. The raised range is cautious arithmetic: with $71.9 million already booked in H1, it implies only $68–78 million for the whole second half, or about $34–39 million a quarter. That is below Q2's $45.7 million. Either management is building in a cushion, or Q2 had timing benefits (for example, UK centers catching up after launch) that it does not expect to repeat. The filing does not say which.

To extend its funding, Autolus borrowed $75 million from Perceptive Advisors on July 30, 2026 under a five-year, interest-only facility of up to $250 million. It can draw another $25 million at its option within six months, and $150 million more depends on hitting revenue milestones. The loan costs one-month SOFR (with a 3.50% floor) plus 7.25%, so at least 10.75% a year, or about $8 million of annual interest on the first $75 million. Perceptive also received warrants to buy up to 3.5 million ADSs at $1.9314. Including the first two tranches ($100 million), management expects cash to last into Q2 2028.

Takeaway: Q2 2026 is the quarter AUCATZYL started making money on each dose: gross margin went from negative to about 55% while sales doubled. That fixes the product's economics, but not yet the company's. A $43.8 million quarterly operating loss, new debt costing at least 10.75% a year, and a raised guide that implies slower sales in H2 mean Autolus still needs sales to keep compounding to reach its own target of cash lasting into 2028 without raising more equity.

What to watch next

• Q3 revenue vs. the implied guide. A third quarter near Q2's $45.7 million would make the $140–150 million range look conservative. A drop toward $35 million would suggest Q2 included one-time catch-up.
• Whether gross margin holds above 50%. One strong quarter after a 6% Q1 is not yet a trend. Management's 65–70% target is the long-run benchmark.
• SG&A without severance. With the layoffs "substantially complete," Q3 SG&A should show whether the $15 million annual savings are real.
• Clinical catalysts: a longer-term lupus (SLE) update from the CARLYSLE trial has been submitted to the ACR meeting in Q4 2026; first data for AUTO8 in light-chain amyloidosis are due by year-end 2026; first progressive MS data (BOBCAT) in Q1 2027; pediatric ALL Phase 2 (CATULUS) data at the end of 2027; and pivotal lupus nephritis (LUMINA) data in 2028. The autoimmune programs are what could eventually take obe-cel beyond a niche leukemia market.