DWTX — Dogwood Therapeutics, Inc.

Company context · prepared 2026-08-13 · company sweep 2026-08-13 · USD · framework v5.8.0

How to read this. Written for a reader who has not studied pharmacology or finance. General terms are defined once in framework/04-glossary.md. Tags: [VERIFIED — source] / [UNVERIFIED] / [WEB ESTIMATE — source, date].

Program analyses for this company are the sub-folders listed in C.2.

C.0 Company-tier coverage — CLEARED

Regulatory tier: not called at all, correctly. 02-connectors.md makes it conditional on a regulatory catalyst, and the only catalyst on this company’s pipeline is a trial readout.

C.1 What the company is

Dogwood Therapeutics is a development-stage biotechnology company in Alpharetta, Georgia, that develops non-opioid medicines for pain and neuropathy. It was formed in October 2024 when Virios Therapeutics (Nasdaq: VIRI) combined with Wex Pharmaceuticals, a subsidiary of the Hong Kong-listed CK Life Sciences Int’l (Holdings) Inc. (HKEX: 0775), and renamed itself [VERIFIED — GlobeNewswire 2024-10-07; company 8-K exhibit 99.1 filed 2026-08-13]. CK Life Sciences remains its largest shareholder [VERIFIED — 8-K exhibit 99.1, 2026-08-13].

It runs four programs and has no revenue of any kind — the income statement in the 2026-08-13 Q2 release shows $0 revenue in every one of the four periods presented [VERIFIED — 8-K exhibit 99.1, 2026-08-13].

It is effectively a single-asset company. Halneuron, the purified tetrodotoxin injection inherited from Wex, is the only program with a catalyst, the only program in active clinical development that the company itself funds, and the subject of the entire “Key Highlights” section of its own quarterly release. The two legacy Virios antiviral combinations, IMC-1 and IMC-2, were signed away in April 2026 to a development and commercialisation partner, with any proceeds flowing partly to holders of the contingent value rights issued in the 2024 merger [VERIFIED — company press release 2026-04-23; 8-K exhibit 99.1 2026-08-13]. The fourth program, SP16, is a Phase 1b asset whose trial is paid for by the National Cancer Institute rather than by Dogwood [VERIFIED — 8-K exhibit 99.1, 2026-08-13].

C.2 Pipeline — every program

Every row returned by fetch_company_drugs on 2026-08-13 appears below, including the rows with no catalyst.

No row on this pipeline is marked Failed. Two rows — IMC-1 and IMC-2 — are functionally shelved rather than failed, which is a distinction worth stating plainly: the company stopped paying for them and gave them away, and that is a track record about capital, not about biology.

One has_catalyst: true row on the whole pipeline. That is the finding the attribution block in the program document rests on: there is no second event on this ticker for a price move around the Halneuron readout to be confused with.

C.3 Financial position

Two other numbers on the same balance sheet matter and do not fit the rows above. Total stockholders’ equity is $68,366,108 against a market capitalisation of about $65M [VERIFIED — 8-K exhibit 99.1, 2026-08-13]. That is not a value signal: total assets of $80,495,480 against $9.6M of cash means roughly $71M of the balance sheet is goodwill and in-process research and development booked when Virios acquired Wex, not cash or anything that can be spent. The company said so itself this quarter — it recognised a $9,190,897 non-cash IPR&D impairment precisely because the market capitalisation had fallen below stockholders’ equity, which management treated as an impairment indicator. An accounting write-down triggered by the share price is a lagging record of the market’s opinion, not new information about the drug.

Dilution history. Every raise in the last 24 months, plus the conversion that reshaped the share count.

Is a raise before the next catalyst likely? The honest answer is that the company would like to avoid one and may not be able to sell shares even if it wanted to. Two facts point the same way. First, its own runway statement says the cash reaches the readout and stops there — not past it. Second, and less obvious, its access to the shelf is close to exhausted. Dogwood is a “baby shelf” issuer: General Instruction I.B.6 of Form S-3 caps what it may sell in any twelve months at one-third of the market value of stock held by non-affiliates. The January 2026 prospectus supplement discloses non-affiliate holdings of 5,075,387 shares, and states that after $4,871,889 of prior sales the company had exactly $6,666,157 of remaining capacity — which that offering then consumed in full, to the dollar [VERIFIED — 424B5 filed 2026-01-13]. At today’s $1.93, one-third of a 5.08M-share non-affiliate float is roughly $3.3M, against $6.67M already sold inside the trailing twelve months [UNVERIFIED — modelled on the verified inputs above]. On that arithmetic the registered shelf, including the Northland at-the-market facility, is effectively unusable until January 2027. What remains is a private placement or a registered direct into a strong tape — which is to say, a raise that is far easier to do after good news than before it. That is the shape of the financing risk here: not a raise that pre-empts the readout, but a company with no comfortable way to survive a bad one.

C.4 Valuation frame

The 52-week range is not comparable across its own span, and saying so is more useful than the number. The $9.50 high printed on 2025-09-29, when 2,293,162 common shares were outstanding; the $1.28 low printed on 2026-04-29, after 24,453,588 shares arrived from the November preferred conversion and 2,338,948 more from the January offering. The daily price series is adjusted for corporate actions but not for issuance, so the high and the low describe two very differently capitalised companies. At the $9.50 high the equity was worth roughly $22M; at the $1.28 low, roughly $43M. The stock is down 80% from its high and the enterprise value is up. Anyone reading “8% of the 52-week range” as room to run should read that sentence first.

C.5 Ownership and flow

Nobody is short this stock and nobody is long it either. The short position is 73,790 shares — under two days of trading — and hf_holding_count is zero, confirmed by an empty fetch_company_hedge_fund_holdings. There is no squeeze to engineer and no crowded long to unwind. What there is instead is an extraordinarily thin float: roughly 5.1M shares in non-affiliate hands, trading an average of $94,762 a day over the last thirty sessions [VERIFIED — lib/prices.mjs over data/prices/DWTX.json]. A hundred thousand dollars of buying is a day’s volume. That cuts both ways and is the single most important structural fact about this equity.

Worth noting on the BPIQ short field: short_float reads 0.01398, i.e. 1.398%, against the FINRA-derived 1.45% of non-affiliate float. On this ticker the two agree, which is a different result from CNTB on the same day, where BPIQ’s figure was 3.4× the FINRA reading under every interpretation. The field’s units remain unconfirmed as a general matter; here it happens to be right.

Insider flow: routine grants only, no conviction either way. fetch_company_insider_transactions returned zero rows across three attempts, which is the documented connector bug and is never evidence of no activity. Reading the EDGAR Form 4 feed directly for the same window finds thirteen filings and not one open-market purchase or sale:

• 2026-06-17 — four director option grants of 16,800 options each (Abel De La Rosa, John C. Thomas, Richard James Whitley, David R. Keefer), transaction code A, dated 2026-06-16.
• 2026-03-06 — four grants including 330,000 options to chief executive Gregory Scott Duncan, code A, dated 2026-03-05.
• 2025-12-05 — four further grants.
• 2025-11-25 — the CK Life Sciences preferred conversion described in C.3.

No insider has bought a share in the open market at any price in this window, and none has sold one either. Against a stock that fell from $9.50 to $1.28 over that period, the absence of purchases is the more informative half.

C.6 Options chain and its readability

There is no readable options chain for this company, and this sweep could not establish whether one exists at all. BPIQ’s fetch_company_options_data returned nothing across three attempts, and the Yahoo cross-check that would normally settle the question was blocked by a throttle before the chain request could be made. Every row below is therefore unfilled for a stated reason rather than left blank. No options-implied move is used anywhere in the program document, and the program’s expected-move reading is given as a bracket built from this ticker’s own history instead.

C.7 Reaction to past catalysts

The table is worth stating in one line, because it is unusual: this ticker’s largest single move on positive drug news was downward. Nothing here supports an assumption that a positive Phase 2b topline is automatically a positive share-price event, and the mechanism that broke the link in December — an unfunded company announcing good news into an imminent raise — has not been removed.

C.8 Company data-quality flags

• market_cap is computed off the previous close. $64,736,884 ÷ the EDGAR-filed 33,629,553 shares is exactly $1.9250, the reported previous close, against a $1.93 last trade. The documented bug, firing cleanly on this ticker.
• enterprise_value is unreconcilable against every available cash figure. $48,459,172 against its own $64,736,884 market cap implies $16,277,712 of cash. The company held $9,610,033 at 2026-06-30 and $13,227,839 at 2026-03-31 — BPIQ’s own stale field. No reading reconciles. Recomputed to $55.45M.
• cash is one quarter stale, timestamped 2026-03-31 by finance_updated_at while the 2026-06-30 figure was public from 08:50 on the morning of this sweep.
• max_52_week_position reproduces the previous-close bug here. 0.07847 is (1.925 − 1.28) ÷ (9.50 − 1.28) to five decimals; the last-price computation gives 0.0796. The field is not read either way (C.4 derives the position from the cache), but this is the second documented outcome for a bug 02-connectors.md records as firing on some tickers and not others.
• short_float agrees with FINRA on this ticker. 0.01398 against 1.45% of non-affiliate float. Recorded because the same field was 3.4× wrong on CNTB eight hours earlier: the field is unreliable, not reliably wrong.
• fetch_company_insider_transactions returned zero rows across three attempts while thirteen Form 4s sit in the EDGAR feed over the same window. Third ticker on which this is now recorded (after TLSA and CNTB). The call completes rather than erroring, so an empty result is indistinguishable from a silent failure from the outside.
• fetch_company_options_data returned no output across three attempts. Not previously recorded in 02-connectors.md for this connector. Because the tool’s contract returns None for both failure and emptiness, this sweep cannot say whether DWTX has no listed options or whether the request failed. The Yahoo cross-check that would have settled it was itself blocked.
• BPIQ’s own Halneuron row is internally inconsistent on the date. catalyst_date_text reads Q3 2026 while the note field on the same row records the company’s 2026-07-20 guidance as “Topline data still expected Fall 2026” — two different periods in one record. The company’s 2026-08-13 release adds a third reading, “the fourth quarter of 2026”. Carried into the program document’s readout block as an unresolved disagreement rather than averaged away.
• has_catalyst is false on the SP16 row (19600) despite a populated catalyst_date. The opposite direction from the stale-true bug documented on CNTB, and the correct answer here: a Phase 1b enrollment start is not a market-moving catalyst. Recorded so that a later reader does not treat the populated date as a second event on this ticker.