ArriVent's ARR-217 shows 44.8% responses, but only 24.1% confirmed; stock falls 28.8%

ArriVent BioPharma, based in Newtown Square, Pennsylvania, is a clinical-stage cancer drug company that has never sold a product.¹ Its lead drug, firmonertinib, is a daily pill for lung cancers driven by certain gene mutations, licensed from Shanghai Allist Pharmaceuticals in June 2021.² It is approved in China but not in the United States.³

On October 6, 2026, ArriVent said FURVENT, a 398-patient trial of firmonertinib against chemotherapy as a first treatment, missed its main goal of delaying tumor growth as judged by outside reviewers.¹ Patients on the 240 mg dose went a median 11.0 months before their cancer grew, against 9.5 months on chemotherapy, and the gap was not statistically firm.² The company is still weighing the full data before it picks a path for the drug, and the stock fell from $28.47 on October 5 to $15.12 on October 6.³

Firmonertinib still has a second large trial, ALPACCA, in patients with a different group of mutations called PACC, and ArriVent says it is still enrolling.¹ An earlier, smaller study in that group showed tumors shrinking in 68% of patients at the 240 mg dose.² Rivals are close in exon 20, the group where FURVENT failed: sunvozertinib won approval in the United States in July 2025 for previously treated patients, and the quarterly report notes a February 27, 2027 decision date for zipalertinib in previously treated exon 20 patients.³

The patent office also agreed in July 2026 to re-examine one firmonertinib patent.

That leaves more weight on ArriVent's antibody-drug conjugates, which are antibodies that carry a chemotherapy payload straight to cancer cells showing a target protein. ARR-217, licensed from Lepu Biopharma on January 21, 2025, aims at cadherin-17, a protein often overexpressed in colorectal and other gut cancers but scarce in normal tissue.¹ It carries an exatecan-based payload, ArriVent holds the rights outside Greater China, and ArriVent dosed its first patient in the study in March 2026.²

In a filing on October 13, 2026, ArriVent said a second conjugate, ARR-002, aimed at ovarian and endometrial cancer, had dosed its first patient in the United States.¹ ArriVent licensed its Greater China rights to Allist for payments of up to $80.6 million.²

The October 6 drop also shows in holder filings. OrbiMed funds sold 151,366 shares at $11.71 that day and still hold 2,875,962 shares, about 5.8% of the company, which they say they may sell.

The law firm Kirby McInerney has announced an investigation on behalf of ArriVent investors, but says no lawsuit has been filed.

The data come from an open-label study in the United States and China that had treated 123 patients with heavily pretreated gut cancers across 6 dose levels from 2 to 8 mg/kg by the August 5, 2026 cutoff.¹ Of those patients, 113 had colorectal cancer.² ArriVent presented the results at the European Society for Medical Oncology congress and put out its release on October 25, 2026.³

At the 6 mg/kg dose, 13 of 29 colorectal patients had a partial response, meaning their tumors shrank by a set amount, for a 44.8% response rate.¹ ArriVent's own quarterly report quotes firmonertinib's earlier result as a confirmed response rate, and just 7 of these 13 are confirmed.² Four more are pending confirmation and 2 are listed as unconfirmed.³

That makes the confirmed rate 24.1%, and even if all 4 pending responses hold, it would be 37.9%.¹ Across all doses of 4 mg/kg or more, 25 of 73 patients responded, a 34.2% rate, but only 15 are confirmed, or 20.5%.² The broader measure is stronger: 27 of 29 patients at 6 mg/kg, or 93.1%, saw their tumors either shrink or stop growing.³

Most responders were still on treatment at the cutoff, so nobody yet knows how long the responses last.

Across all 123 patients, 37, or 30.1%, had a severe or worse side effect tied to the drug, led by low white blood cell counts in 12.2% and diarrhea in 8.9%.¹ At 6 mg/kg the rate was 20.4%, but that figure covers 49 patients, a larger group than the 29 used for the response rate.² No patient died of a drug-related side effect, and no lung inflammation was seen.³

AVBP fell 28.8%, from $21.02 at the October 23 close to $14.96 at 10:36 on October 26, while the S&P 500 fund rose 0.7%.¹ The biotech fund XBI barely moved, at $150.25 against a $149.53 close.²

The stock is now below its $15.12 close of October 6, the day FURVENT failed, so the whole run from $16.69 on October 16 is gone.¹ In our price file, which starts in mid-September, the highest close was $30.335 on September 22.² The release gives no reason for the drop; our reading is that buyers had priced in a cleaner confirmed rate.

At $14.96, the 49,307,729 shares outstanding on August 10, 2026 are worth about $738 million, down from about $1.04 billion at the October 23 close.¹ ArriVent had $373.1 million of cash and short-term investments at June 30, 2026, about $7.57 a share, so the market puts about $365 million on the drugs before about four months of spending since June 30 and $11.0 million of milestones then payable.²

The losses are mostly cash. The net loss was $49.9 million in the second quarter of 2026 and $93.2 million in the first half, which included $12.2 million of stock-based pay that uses no cash.¹ Cash used in operations was $81.5 million in the first half, or roughly $41 million a quarter.²

ArriVent's August 12 results release says the cash funds operations into 2028, and no filing since the October 6 FURVENT miss has repeated that forecast.¹ The quarterly report's formal test only promises at least twelve months from its issue, and it says more capital will be needed.²

ArriVent has no long-term obligations other than an operating lease, and nothing has been drawn on its bank loan.¹ That loan offers up to $75.0 million, of which $50.0 million depends on conditions and milestones, and any draw would pay at least 6.00% and carry a first claim on nearly all assets except intellectual property.² With no debt and no operating profit, debt-to-earnings and interest-cover ratios do not apply.

ArriVent also owed $11.0 million of earned development milestones at June 30, 2026.

New shares are how ArriVent pays its bills. In the first half of 2026 it sold 5,358,318 shares for $140.0 million through an at-the-market program, which sells stock into the market a little at a time, and about $229.1 million of that program was left at June 30.¹ There were also 7,301,993 options at an average exercise price of $18.97, of which 2,220,464 at $11.19 could be exercised, plus 963,469 pre-funded warrants with a near-zero exercise price.²

Counting all of them with the 49.3 million shares gives about 57.6 million, though the options' average exercise price sits above today's price.

The Lepu deal has two sides. For the rights outside Greater China, ArriVent paid $40.0 million up front and owes up to $0.28 billion in development and regulatory milestones, up to $0.89 billion in sales milestones, and royalties from high single digits to low teens.¹ Lepu earned $12.0 million of development milestones in the first half of 2026, of which $6.0 million had been paid by June 30.²

The strongest bull case is that the drug held disease in check in 93.1% of colorectal patients at 6 mg/kg, and most responders were still on treatment, so the confirmed count can still rise.¹ The 6 mg/kg dose, one of two still being tested in dose optimization, looked milder than the study as a whole, with 20.4% severe side effects against 30.1% overall, and no drug-related deaths or lung inflammation.² At about $365 million for everything but June 30 cash, buyers also get a second Phase 3 lung trial that is still enrolling and a runway the company put into 2028 on August 12, before the FURVENT miss.³

The strongest bear case starts with the headline itself: the confirmed rate is 24.1% at 6 mg/kg and 20.5% across doses of 4 mg/kg and up.¹ Next, there is no data yet on how long responses last, and the safety group at 6 mg/kg is not the group used for the response rate.² Last, ArriVent burned $81.5 million in six months, has about $229.1 million of share sales ready, and its lead drug's path is undecided, so new stock sold at low prices is a real risk.³

The company's own goals are plans, not results. It has moved ARR-217 into dose optimization at 5 and 6 mg/kg, expects that data in 2027, and calls the drug potentially best-in-class.

The webinar at 8 am Eastern on October 26 has already taken place, before the 10:36 price we quote, and its content is not in our sources; the dates still ahead are the third-quarter report, for which our sources give no date, the rival's February 27, 2027 decision date, and the dose-optimization data in 2027.¹ The bull case is wrong if the 4 pending responses fail to confirm or responses prove short, which ArriVent has dated only as dose-optimization data in 2027; the cash side is tested at the third-quarter report, whose date is not yet announced (the second-quarter report came on August 12).² The bear case is wrong if the confirmed rate climbs toward 37.9% with lasting responses and the third-quarter report shows cash use near the first-half pace.³

The release gives no median length of response or time before cancer grows for ARR-217, and most responders were still on treatment at the August 5, 2026 cutoff.

The release does not say whether patients were tested or selected for cadherin-17 levels, which matters for how many patients the drug could reach.

ArriVent has not yet reported its cash at the end of the third quarter, or how many shares it sold in that quarter.

The filings do not name the conditions and milestones behind $50.0 million of the $75.0 million bank loan, so whether the FURVENT miss affects access to it is unknown.

The filings give the milestone caps but not which milestone, if any, the Phase 1b start or the 2027 data would trigger.

No transcript or slides from the October 26 webinar, and no analyst notes, are in our sources, so the reason for the drop is our reading, not a reported fact.

The release gives no comparison with standard late-line colorectal treatments or with rival cadherin-17 drugs.