Dianthus Therapeutics Inc. (NASDAQ: DNTH) — Strong Signals, Pivotal Risk Already Capitalized

Published: 2026-09-04 · Verdict: Watch · Entry price: $68 · Price target: $85 · Research confidence: High (87%)

Executive conclusion

Analyst take — WATCH at $107.85; dilution-adjusted central value $85; preferred entry $68 or below. Claseprubart has moved beyond a mechanism-only story: the 65-patient MaGic study produced concordant randomized effects on MG-ADL, QMG, and MGC in generalized myasthenia gravis, while CAPTIVATE reported 30 responders among its first 40 evaluable CIDP patients. Dianthus also held $1.190 billion of cash and investments at June 30, 2026, carried no financial debt, and completed its at-the-market program after quarter-end. These facts materially reduce near-term financing risk. They do not eliminate clinical, commercial, or per-share dilution risk. [S1][S4][S5]

At the September 3 close, 55.934 million reported common shares at $107.85 represented approximately $6.03 billion of common-equity value. Including 1.347 million essentially paid-up pre-funded warrants produces approximately $6.18 billion of economic-basic equity value. Deducting the June cash and investment balance gives economic-basic enterprise value near $4.99 billion. Applying the treasury-stock method to 7.178 million outstanding options at a $32.56 weighted exercise price adds roughly five million net shares and raises effective enterprise value to approximately $5.5 billion. These are analyst calculations rather than company-reported valuation measures; they also mix the September price with the last reported balance sheet and therefore should be read as a range, not false precision. [S1][S18]

Central thesis and variant perception. Claseprubart may combine competitive efficacy with a commercially meaningful administration advantage: selective inhibition of active C1s, preservation of the lectin and alternative complement pathways, an approximately 60-day serum half-life, and intended self-administration every two or four weeks. The balance sheet allows DNTH to test that proposition across gMG, CIDP, and multifocal motor neuropathy without a near-term financing cliff. The variant conclusion is that the market is assigning pivotal-quality probability to evidence that remains partly small, open label, responder enriched, modeled, or contradicted by an important same-target precedent. [S2][S5]

MaGic was positive, but it randomized only 65 patients, treated safety as its primary objective, used one-sided nominal statistics in the initial release, and did not include a randomized 300-mg-every-four-weeks arm. CAPTIVATE’s 75% figure was generated in an open-label induction phase after rapid transition from immunoglobulin; it cannot be interpreted as a placebo-controlled response rate. Its initial cohort comprised approximately 62% standard-of-care-treated, 23% refractory, and 15% treatment-naïve patients, but numerical response rates by stratum have not been disclosed. Only responders proceed to randomized withdrawal. [S4][S5][S13]

The strongest disconfirming evidence is Sanofi’s June 2026 discontinuation of MOBILIZE after an independent committee concluded that Phase 3 riliprubart was unlikely to demonstrate sufficient efficacy in refractory CIDP. Earlier open-label Phase 2 results had reported improvement or stability in most patients, including approximately half of refractory patients improving. This sequence directly challenges the proposition that active-C1s biology or open-label CIDP response is already clinically validated. It does not prove claseprubart will fail: molecules, potency, dosing, populations, and designs differ, and MOBILIZE addressed refractory disease rather than CAPTIVATE’s entire population. [S6][S7]

Principal counter-case. The optimistic case remains credible. MaGic’s effects were directionally consistent across patient-reported and objective measures, the lower 300-mg dose showed a more favorable antinuclear-antibody profile than 600 mg, and CAPTIVATE responders also improved on grip strength, MRC sum score, and I-RODS. If EMERGE reproduces efficacy with both Q2W and Q4W dosing, CAPTIVATE demonstrates durable relapse prevention, longer exposure remains free of clinically important drug-induced lupus or disproportionate serious infection, and the eventual label is less burdensome than terminal-complement alternatives, claseprubart could support a large multi-indication franchise. [S4][S5][S8][S9]

Conviction: moderate. The asset evidence is too strong for a reflexive short thesis, but the valuation provides limited protection against ordinary late-stage attrition. An illustrative rNPV produces approximately $16, $85, and $216 per future diluted share in bear, base, and bull cases. That dispersion is real rather than cosmetic: modest changes in indication-level success probabilities, launch timing, peak sales, label burden, patent duration, and diluted shares change value by billions of dollars.

Evidence that would change the call. The view would improve if MoMeNtum shows a clear, clinically coherent delay in immunoglobulin rescue; CAPTIVATE Part B demonstrates a large and durable relapse-prevention effect with a commercially relevant screening-to-randomization yield; EMERGE validates Q4W dosing; or DNTH212 demonstrates both intended pharmacodynamic mechanisms at a tolerable chronic dose. It would deteriorate if MMN objective measures fail to corroborate the headline, CAPTIVATE resembles the riliprubart sequence, Q4W is materially weaker than Q2W, persistent high-titer ANA progresses to clinical autoimmunity, serious infection or monitoring burdens emerge, or diluted shares grow faster than risk-adjusted asset value. [S1][S5][S6][S14][S15]

Stock Price Action — Five-Year Event Map

The five-year chart contains a structural discontinuity. Prices before September 2023 represent Magenta Therapeutics, a transplant-conditioning developer whose programs and risk profile were unrelated to current Dianthus. The series is reverse-split adjusted, but adjustment cannot make the predecessor’s biology, capitalization, or operating history comparable. The economically relevant DNTH history begins with the September 2023 reverse merger. [S25][S30]

The split-adjusted five-year series ranged from the mid-single digits during Magenta’s collapse to above $130 in 2021. DNTH closed at $107.85 on September 3, 2026. Its recent 52-week range was approximately $24.26 to $117.88, placing the current price near the upper end after a more than fourfold move from the low. Price history is factual; the causal interpretation of each move is necessarily an inference unless tied to a contemporaneous disclosure. [S18]

• 2021–early 2023: predecessor impairment. Magenta’s adjusted price fell from above $100 to single digits. In December 2022 it paused MGTA-117 dosing after serious adverse events; a January 2023 update disclosed a patient death considered potentially related to treatment. This is not adverse evidence about claseprubart. It is, however, a relevant demonstration of how quickly a clinical-stage enterprise can converge toward residual cash when safety invalidates its principal program. [S31]

• May–September 2023: business discontinuity and recapitalization. Magenta agreed to combine with private Dianthus, the transaction closed in September, the shares underwent a 1-for-16 reverse split, and the ticker changed to DNTH. The transaction combined Magenta’s remaining cash, Dianthus’ programs, and concurrent private financing. Pre-merger return charts therefore should not be used as an own-history valuation multiple for the present company. [S26][S30][S36]

• January 2024: balance-sheet validation. DNTH sold approximately 14.5 million common shares and 4.7 million pre-funded warrants at a $12 common-equivalent price, raising approximately $230 million gross. The financing materially extended runway but also expanded the economic share base by roughly 19.2 million shares. The positive price response can reasonably be attributed to lower financing risk, although specialist participation does not independently validate clinical efficacy. [S27]

• September 8–11, 2025: first randomized efficacy re-rating. After MaGic top-line results, the close moved from $26.50 on September 5 to $31.80 on September 8 and $35.57 on September 9. DNTH then issued common shares and pre-funded warrants at a $33 equivalent price for $288.4 million gross. This was the first major clinical de-risking of the current business and another example of financing after appreciation rather than during distress. [S4][S18][S28]

• March 9–12, 2026: CIDP optionality and transformative financing. The early CAPTIVATE continuation decision moved the stock from $65.20 to $79.23 and then $86.92. DNTH subsequently sold approximately 8.47 million common shares and 0.40 million pre-funded warrants at an $81 equivalent price for $719 million gross. The balance-sheet improvement was real; the clinical inference remained preliminary because the continuation decision came from open-label Part A responders, not randomized relapse prevention. [S5][S18][S29]

• June 10–12, 2026: class read-through tested in the market. DNTH fell from $84.57 on June 9 to $76.42 on June 10 when Sanofi announced the MOBILIZE discontinuation, then rose to $82.07 on June 12 after DNTH presented detailed CAPTIVATE evidence. The sequence suggests investors treated riliprubart as relevant but not dispositive. That is a market inference, not clinical proof of molecule-specific superiority. [S5][S6][S18]

• August–September 2026: premium expectations. Following the Q2 update and DNTH312 disclosure, DNTH reached an intraday high near $117.88 on August 19 and closed at $107.85 on September 3. By this point the market was assigning substantial value to multi-indication success despite the absence of randomized CIDP results, pivotal gMG results, or human DNTH212/DNTH312 efficacy evidence. [S1][S16][S18]

Verdict: The post-merger appreciation reflects genuine changes—controlled gMG activity, broader clinical optionality, and removal of near-term financing pressure. It also moved the security from cash-supported optionality toward a valuation that requires execution. The predecessor collapse is not comparable operating evidence, but it remains an honest reminder that abundant cash does not prevent permanent impairment when the key biological thesis fails.

Business Overview

Dianthus is a precommercial biotechnology company developing biologics for severe autoimmune diseases. It has no approved product, no product sales, no established commercial infrastructure, and no demonstrated ability to manufacture at launch scale or secure reimbursement. Its present economic output is not revenue growth; it is the change in probability, timing, protected life, and eventual profitability of future products. License and collaboration revenue is episodic and immaterial relative to research spending. [S1][S2]

Claseprubart

Claseprubart, formerly DNTH103, is a fully human IgG4 monoclonal antibody designed to bind selectively to the active form of complement protein C1s while avoiding inactive pro-C1s. C1s is part of the classical complement pathway. The proposed mechanism is to interrupt antibody-driven classical-complement injury while preserving alternative- and lectin-pathway activity. The molecule includes a YTE Fc modification intended to extend exposure; a Phase 1 study in 60 healthy volunteers produced a company-reported serum half-life of approximately 60 days. DNTH is developing a pre-filled device for short subcutaneous self-administration every two or four weeks. [S2][S5]

The customer proposition has three potential components. First, inhibition of classical-complement activity may reduce tissue injury in antibody-mediated neuromuscular disease. Second, pathway selectivity could preserve more immune function than terminal C5 blockade. Third, infrequent at-home administration could reduce infusion-center dependence and improve persistence. None is yet a commercial fact. There is no head-to-head efficacy trial, no approved label, no payer contract, and no evidence that Q4W will retain the same clinical effect as Q2W across weight and disease-severity strata.

The same pharmacology creates two-sided risk. A long half-life supports convenience and smoother trough exposure, but it also slows clearance after an unexpected infection, autoimmune complication, hypersensitivity reaction, or other delayed adverse event. DNTH itself identifies prolonged side effects and difficulty switching therapies as possible consequences of extended exposure. Convenience should therefore be valued alongside reversibility, not independently of it. [S2]

Claseprubart is being developed in three neuromuscular indications. EMERGE is a roughly 195-patient Phase 3 gMG study comparing 300 mg Q2W, 300 mg Q4W, and placebo in acetylcholine-receptor-antibody-positive patients. CAPTIVATE is a Phase 3 CIDP program with open-label induction followed by randomized withdrawal among responders. MoMeNtum is a 46-patient Phase 2 MMN trial evaluating 300 mg, 600 mg, and placebo; its primary focus is safety, and company materials state that it is not powered for conventional statistical proof of efficacy. [S1][S12][S13][S14]

Using one molecule across three indications can create operating leverage. Manufacturing development, device engineering, toxicology, safety databases, regulatory expertise, medical affairs, and neurologist relationships can be reused. If two or more indications succeed, marginal commercialization economics may be attractive. The same architecture creates concentration: a molecule-level safety problem, manufacturing failure, adverse patent result, or invalidated target thesis can impair the entire franchise at once. The popular phrase pipeline in a product is thus both an upside mechanism and a correlated-risk warning.

DNTH212

DNTH212, also known as LBL-047, is licensed from Nanjing Leads Biolabs. It combines BDCA2 engagement on plasmacytoid dendritic cells, intended to suppress type-I-interferon production, with BAFF/APRIL inhibition intended to reduce B-cell and antibody activity. Dianthus controls rights outside Greater China, while the partner is conducting the initial China Phase 1 program. The registered study begins in healthy volunteers and includes a later single-dose systemic-lupus-erythematosus portion. Management expects an initial healthy-volunteer pharmacology update by year-end 2026, while the registered study continues into 2027. An interim pharmacodynamic disclosure should not be mistaken for patient proof of concept. [S1][S15]

The license has substantial contingent economics. DNTH had paid $30 million by June 2026 and may owe another $8 million when a Dianthus-led Phase 1 trial begins. It could owe up to $962 million across development, regulatory, and commercial milestones covering several geographies and indications, plus mid-single-digit to low-double-digit royalties. These are not balance-sheet debt and are payable principally on success. They nevertheless reduce shareholder participation in gross product economics and belong in any DNTH212 rNPV. [S1]

The dual mechanism could be valuable if both pathways contribute to the same disease and can be modulated without excessive immunosuppression. It could also create dose-selection difficulty: the dose needed for meaningful pDC/type-I-interferon suppression may not match the dose that safely affects BAFF/APRIL biology. The correct near-term evidence is simultaneous, dose-responsive engagement of both intended mechanisms, usable pharmacokinetics, and acceptable immunoglobulin and infection effects—not a generic statement that each component pathway is validated elsewhere.

DNTH312

DNTH312 is an internally developed bifunctional candidate combining claseprubart-like C1s inhibition with a TACI domain targeting BAFF/APRIL. The disclosed dataset is preclinical, including nonhuman-primate pharmacokinetics and immunoglobulin observations. Management targets Phase 1 readiness by year-end 2027. Comparisons with published povetacicept data are cross-study and preclinical; they cannot establish superior human exposure, safety, or efficacy. [S16]

The program may extend internal complement expertise and create another route into autoantibody-driven disease. It may also compound immune suppression and duplicate mechanisms already represented in the portfolio. With no human dose, pharmacodynamic, immunogenicity, or safety evidence, DNTH312 should receive little central valuation today. Its economic value is better treated as a contingent option whose probability should rise only after candidate nomination milestones and human data.

Revenue architecture and segments

Reported revenue has come from licenses and collaborations, chiefly regional rights, rather than end-customer demand. Revenue was $1.5 million in 2021, $6.4 million in 2022, $2.8 million in 2023, $6.2 million in 2024, $2.0 million in 2025, and $1.2 million in the first half of 2026. Cumulative revenue over that period was only about $20.2 million against approximately $386.8 million of R&D. Revenue multiples are therefore economically meaningless. [S1][S19][S25][S26][S36]

DNTH does not report conventional operating segments because no commercial product business exists. The Tenacia agreement allocates Greater China claseprubart rights and provides upfront, milestone, reimbursement, and royalty economics. It shares regional development cost and regulatory complexity but gives away part of future economics. This is useful non-dilutive funding, not a recurring platform-revenue model. Deferred-revenue timing is too small to explain the balance sheet; equity issuance created the liquidity.

If claseprubart is approved, biologic gross margins could be attractive, but present expenses omit much of the mature cost structure. Commercialization would add launch inventory, manufacturing validation, quality systems, medical affairs, sales personnel, rebates, payer contracting, pharmacovigilance, distribution, device support, and working capital. The current asset-light structure should not be extrapolated into a mature margin without those costs.

Intellectual property and operating footprint

As of February 10, 2026, DNTH disclosed two issued U.S. patents in its selective-C1s-antibody portfolio, plus pending U.S. and foreign applications. The same filing states that one issued patent covers claseprubart and that relevant protection is expected to expire no earlier than 2043, before possible disclaimers or extensions. The statements are not necessarily contradictory—the second issued patent may cover adjacent portfolio claims—but they show why a single 2043 date is insufficient. Composition coverage, active-state-binding claims, methods, formulation, device claims, geography, validity, enforceability, and freedom to operate must all be tested. [S2]

DNTH reported approximately 92 employees, including 66 in R&D, and relies heavily on CROs, CDMOs, clinical sites, technology licensors, and device vendors. Physical capital expenditure is negligible and the company owns no commercial-scale manufacturing network. This limits fixed capital but introduces external quality, capacity, inspection, geopolitical, and continuity risks. The 2025 10-K also identifies reliance on foreign service providers, including entities affected by U.S. biotechnology-supply-chain policy. [S2]

There are no hidden conventional segments, material real-estate assets, dividends, repurchases, or recurring service revenues to provide downside support. The potentially unrecognized assets are scientific know-how, regulatory interactions, manufacturing methods, data, and patent applications. The off-balance-sheet economic claims include trial contracts, manufacturing commitments, success-based license payments, royalties, and the cost of building commercial capability.

Verdict: DNTH is a concentrated research enterprise, not a diversified operating company. Claseprubart has a coherent biological and convenience proposition; DNTH212 and DNTH312 create real but early optionality. The potential economics improve sharply after approval because infrastructure can be reused, but no product revenue, commercial moat, or realized return on research capital exists today. Molecule-level leverage creates the same correlation on the downside that it creates on the upside.

Industry Dynamics

Dianthus competes in autoimmune biologics, where the profit pool can be large despite relatively small patient populations. Chronic treatment, high annual prices, specialist prescribing, and difficult disease biology support attractive gross economics after approval. Those economics attract extensive capital. Scientific scarcity may persist, but commercial scarcity can disappear as several mechanisms reach the same patient and payers gain alternatives.

Generalized myasthenia gravis

gMG already has multiple targeted classes. Approved alternatives include C5 inhibitors such as eculizumab, ravulizumab, and zilucoplan; FcRn therapies such as efgartigimod, rozanolixizumab, and nipocalimab; and CD19-directed therapy. Conventional corticosteroids and other immunosuppressants remain relevant. Products differ in antibody restrictions, onset, durability, dosing, cycling, infection management, route, payer positioning, and physician familiarity. [S2][S8][S10]

The addressable profit pool is consequently smaller than diagnosed patients multiplied by list price. Eligibility, disease severity, prior treatment, antibody status, physician sequencing, rebates, adherence, and discontinuation all divide the market. A new entrant must persuade clinicians and payers to switch from an approved therapy with established reimbursement. Placebo-controlled efficacy is necessary but may be insufficient if the administration profile or label is ordinary.

Claseprubart’s 300-mg MaGic arm produced placebo-adjusted week-13 changes of approximately 1.8 points on MG-ADL and 2.4 points on QMG. Nipocalimab’s FDA summary reports a roughly 1.5-point MG-ADL and 2.8-point QMG placebo-adjusted effect in a different trial. The values establish that claseprubart’s signal is within a clinically credible range; they do not demonstrate cross-trial superiority because baseline characteristics, visit schedules, concomitant treatment, estimands, and statistical hierarchies differ. [S4][S10]

Administration could be economically meaningful. Patients may prefer predictable maintenance and infrequent self-injection over infusion-center visits, daily injection, or treatment cycles. A Q4W label could improve adherence, reduce provider burden, and support share at similar efficacy. The evidence gap is direct: MaGic randomized Q2W regimens, while Q4W support comes from pharmacokinetic modeling and open-label extension exposure. EMERGE must show that trough exposure and clinical benefit remain adequate, especially in heavier or more severe patients. [S5][S12]

CIDP

CIDP has historically depended on intravenous or subcutaneous immunoglobulin, corticosteroids, and plasma exchange. Immunoglobulin can be effective but imposes infusion time, access, headache, thrombosis, supply, and payer burdens. Weekly subcutaneous efgartigimod alfa/hyaluronidase is now an approved targeted alternative. Its registrational program used open-label response identification followed by randomized withdrawal, illustrating both the regulatory usefulness and the interpretive limits of responder-enrichment designs. [S11]

The supply-side opportunity is attractive: replace plasma-derived immunoglobulin with scalable recombinant manufacturing and less frequent self-administration. Yet the same opportunity has drawn FcRn, complement, and other immune-modulating competitors. By launch, DNTH may face a market shaped by several approved targeted products, negotiated rebates, treatment sequencing, and physician experience accumulated before claseprubart’s label.

Riliprubart provides the most relevant external evidence because it also targets active C1s. Sanofi’s Phase 2 program reported improvement or stability across standard-of-care-treated, refractory, and treatment-naïve cohorts. In June 2026, the independent monitoring committee for Phase 3 MOBILIZE concluded that the refractory trial was unlikely to demonstrate sufficient efficacy, and Sanofi discontinued it without identifying a new safety signal. [S6][S7]

The correct read-through is conditional. It reduces confidence that active-C1s inhibition alone guarantees clinical efficacy and shows that uncontrolled improvement can fail to survive late-stage testing. It is strongest for refractory CIDP, the population studied in MOBILIZE. It is weaker for CAPTIVATE’s standard-of-care-treated and treatment-naïve groups, and it does not resolve differences in potency, exposure, or design. Sanofi’s separate treated-patient VITALIZE program also means the class question is not fully closed. Treating MOBILIZE as irrelevant would be promotional; treating it as proof of claseprubart failure would be analytically careless.

DNTH’s in-vitro comparisons report stronger binding and complement inhibition than a riliprubart molecule reconstructed from public sequence information. The assays support a plausible explanation for molecule-level difference, but the comparator was reconstructed and the experiments were sponsor run. Laboratory potency can inform prior probability; it cannot substitute for clinical randomization. [S5]

MMN

MMN is smaller and more concentrated. Immunoglobulin remains the central treatment, and unmet need centers on durable motor control with less treatment burden. DNTH estimates more than 10,000 U.S. patients, but this is a management market estimate rather than an independently audited count. Diagnosis can be difficult, prevalence estimates vary, and only part of the population will be eligible for a new premium biologic. [S2][S17]

Argenx’s empasiprubart, an anti-C2 antibody, is the principal near-term competitor and external mechanism test. Phase 2 data suggested a lower need for IVIg rescue, but estimates were imprecise in small cohorts. The Phase 3 EMPASSION study compares empasiprubart with IVIg in approximately 154 patients. A positive result would validate proximal-complement intervention while potentially giving an established commercial rival a timing and infrastructure advantage. A negative result could weaken the broader complement thesis before DNTH’s program matures. [S34][S35]

MoMeNtum’s December 2026 update can create meaningful option value, but the trial’s small size and safety-primary design matter. Enrollment above the initial target improves information content but does not transform it into a registrational efficacy trial. A persuasive signal needs concordance among time to retreatment, grip strength, MRC score, and MMN-RODS, with transparent handling of rescue rules and missing data. [S1][S14]

Regulation, reimbursement, geography, and barriers

Claseprubart received U.S. orphan-drug designation in MG. Orphan status can provide development incentives and potential exclusivity after approval; it is not an efficacy endorsement and does not prevent competition from clinically different products. Patent duration may be more important than orphan exclusivity if the relevant claims remain valid through the early 2040s. [S1][S2]

The meaningful entry barriers are accumulated human evidence, biologic-manufacturing consistency, regulatory experience, pharmacovigilance, device reliability, specialist relationships, payer access, and enforceable intellectual property. Scientific novelty is only one layer. Switching costs emerge after a patient is stable, a clinician is comfortable, and reimbursement is established; those installed-base advantages currently belong to approved incumbents.

Payers can capture part of the convenience benefit. If several mechanisms achieve similar outcomes, rebates and prior-authorization sequencing can compress net price even while list prices remain high. A therapy replacing immunoglobulin may support a compelling health-economic case, but DNTH will need evidence on total treatment burden, rescue use, persistence, and resource consumption. None is currently disclosed at commercial scale.

The development footprint is international, with Greater China partnerships reducing local cost and regulatory burden while surrendering regional economics. Foreign-exchange exposure is not reported as a meaningful segment, but overseas CRO, CDMO, cell-line, and trial-site dependencies can create cost and continuity risk. U.S. restrictions affecting specified biotechnology providers could require technology transfer, supplier replacement, validation, or delay. [S1][S2]

Capital-cycle lens

Autoimmune biologics exhibit a classic attractive-profit-pool capital cycle. High prices and chronic use attract large-cap pharmaceutical companies, venture-backed developers, platform companies, and mechanism variants. Capital funds more trials and better devices, but also raises the clinical standard and reduces scarcity by launch. Competitor failure can remove capacity, yet it can also invalidate part of a mechanism thesis. DNTH’s balance sheet allows it to stay in the cycle; it does not protect it from the returns compression caused by many credible alternatives.

Verdict: The end markets can support a valuable franchise, especially where an infrequent recombinant biologic replaces immunoglobulin or burdensome complement therapy. Industry profitability after approval is attractive, but competition is already intense in gMG and increasingly so in CIDP and MMN. The Sanofi result raises the standard for active-C1s evidence. DNTH must establish molecule-specific efficacy, dosing, safety, and reimbursement advantages rather than rely on class logic or market-size slides.

Competitive Position

The relevant moat question is not whether claseprubart is scientifically differentiated. It is whether that differentiation can produce durable prescribing preference, favorable net pricing, persistence, and returns after competitors respond. DNTH currently has potential advantages but no demonstrated economic moat because it has no approved product, installed patient base, payer access, or commercial cash flow.

Active-C1s selectivity

Claseprubart binds active C1s while sparing inactive pro-C1s and preserving lectin- and alternative-pathway activity. Relative to C5 inhibitors, this could retain more antibacterial complement function and reduce vaccination, boxed-warning, or REMS burden. Relative to FcRn inhibition, it could address complement-mediated tissue injury without broadly lowering circulating IgG. The mechanism is coherent and supported by pharmacodynamic evidence; commercial superiority remains hypothetical. [S2][S5]

Existing labels provide both validation and caution. Zilucoplan’s gMG label contains a boxed warning for serious meningococcal infection, vaccination requirements, and a REMS. Sutimlimab proves that C1s can be inhibited in humans, but its label warns about serious infections, recommends vaccination against encapsulated bacteria, and discusses autoimmune-disease risk associated with classical-pathway deficiency. These labels do not predetermine claseprubart’s label because molecule, indication, exposure, and data differ. They do refute any assumption that proximal selectivity automatically eliminates infection or autoimmune scrutiny. [S8][S9]

MaGic produced a dose-associated laboratory signal. Newly positive high-titer ANA at any post-baseline assessment occurred in none of 22 placebo patients, one of 17 evaluable 300-mg patients, and eight of 22 600-mg patients. No clinical drug-induced lupus or spontaneous SLE occurred, and some titers declined or reverted. The absence of clinical disease is favorable; the numerical dose pattern suggests real pharmacology rather than random noise. [S5]

FDA agreed during development to remove ANA screening exclusion and routine ANA monitoring and to frame the hypothetical concern as drug-induced lupus rather than spontaneous SLE. That decision should reduce enrollment friction and monitoring burden. It is not equivalent to a final label decision or proof that rare, delayed autoimmunity cannot occur. Dozens of patients over months cannot establish superiority on a rare safety event that might require hundreds or thousands of patient-years to characterize. [S1][S5]

Dosing and device

The approximately 60-day half-life supports the intended Q2W and Q4W product profile. A short, self-administered injection could reduce infusion-center time and improve persistence. If efficacy and safety are otherwise similar, convenience can become an economically durable differentiator because chronic patients and clinicians value predictable control with fewer administration events. [S2][S5]

The Q4W evidence is not yet sufficient. Phase 2 randomized 300-mg and 600-mg Q2W regimens. The proposed Q4W regimen is supported primarily by modeled exposure and open-label observations in patients switched into extension treatment. EMERGE includes both Q2W and Q4W arms and is therefore the first direct randomized test of the proposed interval. A Q2W-only label could still be competitive, but it would narrow the differentiation against weekly and cyclic alternatives.

The device is an execution dependency as well as an advantage. External device technology, fill-finish, human-factors testing, supply continuity, and inspection all matter. A one-click device can strengthen the product profile, but it is unlikely to prevent a well-capitalized competitor from developing another patient-friendly presentation. The moat exists only if the device improves persistence, market share, or net price after accounting for its development and supply cost.

Efficacy evidence

MaGic’s efficacy is the strongest positive evidence. At week 13, placebo improved approximately 2.8 MG-ADL points, compared with 4.6 for 300 mg and 5.4 for 600 mg, producing adjusted differences of 1.8 and 2.6 points. QMG differences were approximately 2.4 and 2.5 points; MGC differences exceeded five points. Directional agreement across patient-reported and objective measures reduces the probability that the entire result was a scoring artifact. [S4]

The statistical and design limitations remain material. Each arm included only 21 or 22 patients. The initial release reported one-sided nominal p-values and treated p below 0.10 as nominally significant. Later two-sided values for the principal MG-ADL and QMG comparisons remained below 0.05, so it would be wrong to claim that the core effects disappear under two-sided testing. The proper concern is small-sample uncertainty, multiplicity, material placebo improvement, and inconsistent secondary or dose-response evidence. [S4][S5]

Minimal symptom expression occurred in approximately 37% of 300-mg patients versus 14% of placebo, but only 27% at 600 mg. Quality-of-life evidence was also less consistent in the higher-dose arm. The absence of monotonic clinical benefit does not invalidate activity; it weakens arguments that more exposure inevitably produces more efficacy and reinforces the importance of selecting a dose that balances benefit and ANA risk.

CAPTIVATE broadens the evidence but remains uncontrolled. Patients switched from immunoglobulin after seven days without a required deterioration period. During open-label Part A, response required at least a one-point adjusted-INCAT improvement, and only responders proceeded to Part B. Among the first 40 evaluable patients, 30 responded, five relapsed, and five remained stable. Responders also improved on grip strength, MRC, and I-RODS, while nonresponders worsened. [S5][S13]

The objective measures make a purely subjective-score explanation less likely. They do not create a concurrent control. Residual immunoglobulin would not mechanically explain every improvement from the Part A baseline, but rapid transition can affect disease trajectory and complicate attribution; regression to the mean, expectation, assessor effects, and selected enrollment remain. The full screening funnel and response rates by prior-treatment stratum are necessary to assess transportability.

CAPTIVATE’s design was modified from a larger three-arm randomized phase that included 600 mg to a smaller two-arm comparison of 300 mg Q2W with placebo. Removing the higher dose may reduce ANA exposure and accelerate execution, but it also eliminates randomized dose-response information in CIDP. Responder enrichment can improve power to demonstrate maintenance while narrowing the result to patients who first show benefit. The commercial question is therefore both whether continuation prevents relapse and how many real-world candidates reach that point.

The riliprubart precedent is the most important challenge. Favorable open-label Phase 2 results in refractory patients did not translate into sufficient efficacy in MOBILIZE. Claseprubart might succeed because of potency, exposure, molecule characteristics, patient mix, or trial design. Until randomized CAPTIVATE results exist, the evidence supports promise—not a validated class advantage. [S6][S7]

Organizational and portfolio position

DNTH’s large cash balance permits parallel trials, manufacturing work, and device development that smaller biotechnology peers might need to sequence. Focus can improve decision speed and concentrate expertise. Shared neuromuscular infrastructure could lower marginal commercialization cost across gMG, CIDP, and MMN.

The company remains disadvantaged against argenx, AstraZeneca, UCB, Johnson & Johnson, Sanofi, and other established competitors in payer contracts, safety databases, field forces, medical affairs, manufacturing redundancy, and physician relationships. Large competitors can bundle contracting, fund head-to-head trials, and tolerate one failed indication. DNTH’s financial capacity is strong for its stage, but its biological concentration is greater.

DNTH212 and DNTH312 are scientific adjacencies, not current moats. Other companies may validate or raise the clinical standard for BAFF/APRIL, BDCA2, and related mechanisms before DNTH reaches patient proof of concept. Dual targeting could deepen response or could add immunosuppression without additive efficacy. Human data must decide.

Durability and financial test

Potential patent protection into 2043 could provide a long commercial period if launch occurs around 2029–2031. Pending claims, methods, formulations, devices, biologic manufacturing, and regulatory exclusivity can add layers. The portfolio is still early, and one disclosed issued patent directly covering claseprubart must withstand validity, scope, and design-around questions. [S2]

The financial proof of moat is not approval alone. It is sustained share at an attractive net price, low discontinuation, manageable selling expense, acceptable manufacturing cost, and incremental returns above the cumulative research and commercialization capital. Without those outcomes, scientific differentiation may create a product but not a durable economic advantage.

Verdict: Claseprubart may possess a differentiated product profile, especially in administration and pathway selectivity. Best-in-disease remains a management aspiration based on cross-trial and preclinical comparisons, not a demonstrated fact. Pivotal replication, Q4W efficacy, a genuinely lighter safety label, and adoption against entrenched products are all required before the potential moat becomes an economic moat. The strongest positive evidence is concordant randomized gMG activity; the strongest negative evidence is the same-target CIDP failure combined with unresolved ANA and label risk.

Growth History and Forward Opportunities

DNTH’s historical growth is measured in clinical maturity and funded option value rather than recurring revenue. License revenue has been irregular and immaterial. R&D rose from $12.6 million in 2021 to $29.4 million in 2022, $32.8 million in 2023, $83.1 million in 2024, and $145.6 million in 2025. First-half 2026 R&D was another $83.2 million. The step-up reflects progression from early pharmacology into multiple Phase 2 and Phase 3 programs. [S1][S19][S25][S26][S36]

MaGic was the pivotal historical milestone. Before September 2025, claseprubart’s value rested largely on mechanism, pharmacokinetics, and healthy-volunteer evidence. MaGic added randomized patient efficacy and justified Phase 3 investment. CAPTIVATE’s early Part A response subsequently added CIDP optionality, although at a lower evidence tier. These events transformed DNTH from a pre-proof-of-concept developer into a multi-indication late-stage company. [S4][S5]

Near-term catalysts

MoMeNtum is the nearest company-specific binary catalyst. Enrollment reached 46 patients, above the initial 36-patient objective, and top-line results are guided for December 2026. The primary endpoint is safety, and the study was not designed as a conventional statistically powered efficacy trial. Time to immunoglobulin retreatment or deterioration, grip strength, MRC score, and MMN-RODS are the decision-useful efficacy measures. [S1][S14]

A strong result would show a visible separation in rescue timing accompanied by objective and functional improvement, consistent across doses and not driven by an unusual rescue decision or post hoc subgroup. A result described only as well tolerated, or one with favorable time-to-rescue but no corroborating motor evidence, would add much less value. Competitor empasiprubart data may change the market standard around the same time. [S34][S35]

DNTH212 healthy-volunteer results expected by year-end 2026 could create early option value. The useful observations are dose-responsive BDCA2/pDC and type-I-interferon pharmacology, evidence of BAFF/APRIL activity at the same dose, half-life compatible with chronic use, controlled immunoglobulin effects, and no prohibitive cytokine, infection, infusion, or immunogenicity signal. Because the registered program continues into 2027 and includes only a later single-dose SLE cohort, this update cannot establish patient efficacy. [S1][S15]

For CAPTIVATE, the next operational catalyst is guidance on Part B timing and progression. The company has guided to provide updated timing by year-end 2026, not necessarily to report randomized data then. The value-determining result is time to relapse or deterioration on continued claseprubart versus placebo, together with absolute relapse rates, screening-to-randomization yield, discontinuations, and subgroup consistency. [S1][S13]

EMERGE is the central medium-term catalyst. Management says dosing began in June 2026 and guides to top-line results in the second half of 2028. A contemporaneous registry update had lagging recruitment fields and later estimated dates. Registry lag is common, but enrollment progress should be monitored because a one-year delay meaningfully reduces present value and increases development spending. [S1][S12]

Longer-duration opportunities

If claseprubart succeeds in gMG and CIDP, MMN could reuse the same manufacturing, device, medical-affairs, and neurologist infrastructure. Other complement-dependent diseases may create additional indications. Those opportunities should not be capitalized simply because the mechanism is broad; each disease requires target relevance, dose selection, controlled efficacy, and a commercial path.

DNTH212 could expand the company beyond neuromuscular disease into SLE, Sjögren disease, dermatomyositis, and related disorders. Management’s combined population and future-market figures are useful opportunity ceilings, not sales forecasts. Diagnosis, eligibility, competitive sequencing, price, persistence, geography, and partner economics can reduce realizable sales substantially. [S17]

DNTH312 may become valuable if both claseprubart and BAFF/APRIL biology are independently validated and a single bifunctional molecule improves efficacy or convenience. The program should remain near option value until human pharmacology establishes that both functions remain active and tolerable in one construct. Advancing it before pivotal claseprubart resolution may be scientifically rational but would increase burn and correlation.

Commercial strategy and geography

Tenacia and Leads provide local Greater China capabilities and cost sharing while reducing retained economics. Outside Greater China, DNTH has not demonstrated whether it will commercialize alone, partner internationally, or license before approval. U.S.-only commercialization might preserve margin but require a sales and payer build-out; broader self-commercialization increases capital and execution risk. A partnership reduces retained economics but can import infrastructure and geographic expertise.

The company’s large cash reserve creates strategic flexibility. It also creates temptation to expand the pipeline before the lead asset is fully validated. The appropriate growth metric is not number of programs or total addressable market. It is probability-adjusted value added per diluted share and per dollar consumed.

Verdict: Clinical progress has been rapid and well funded, but growth remains binary rather than recurring. MoMeNtum and DNTH212 can alter near-term option value; CAPTIVATE and EMERGE remain the much larger determinants. DNTH’s opportunity set is broad enough to support substantial upside, but only controlled indication-level evidence and a credible commercialization plan justify converting market-size claims into valuation.

Financial Quality

A precommercial biotechnology company cannot be evaluated primarily through revenue growth, EBITDA, or near-term EPS. The relevant questions are whether spending produces independent evidence, whether cash reaches decisive milestones, whether accounting obscures dilution or commitments, and whether probability-adjusted asset value grows faster than capital consumed.

Reconciled operating history

The 2021–2022 figures are private Dianthus predecessor results rather than legacy Magenta operations. The subsequent filings treat the merger as a reverse recapitalization, so a mechanically assembled ticker history can combine different entities unless that distinction is preserved. The figures above reconcile to the predecessor statements, annual filings, June 2026 filing, and Company Financials cross-check. [S1][S19][S25][S26][S36]

Cumulative license revenue of approximately $20.2 million equaled only about 5% of cumulative R&D. Cumulative operating cash use was roughly $361 million, while the accumulated deficit reached approximately $428 million by June 2026. No research investment has yet produced commercial cash flow.

Spending accelerated as programs advanced. Operating cash burn rose approximately 112% in 2024 and 65% in 2025. H1 2026 burn of $77.8 million was about 51% above the prior-year period. Q1 burn was $28.9 million and Q2 burn $49.0 million, illustrating uneven trial and manufacturing payments. Annualizing H1 gives approximately $156 million; annualizing Q2 gives approximately $196 million. Neither is a forecast because Phase 3 enrollment, manufacturing campaigns, device validation, headcount, and commercial preparation can raise or shift future cash use. [S1][S19]

Quarterly cost quality

Q2 2026 license revenue was only $0.8 million. R&D increased to $48.7 million from $26.3 million a year earlier, G&A rose to $13.6 million from $8.9 million, and net loss widened to $50.2 million from $31.6 million. Approximately $15.1 million of the R&D increase reflected external spending, including roughly $13 million for claseprubart and $2.1 million for discovery programs; internal cost accounted for the balance. This is consistent with advancing trials rather than a disclosed one-time cost overrun. [S1]

External spending is more variable but not fully flexible. CRO enrollment, CDMO manufacturing slots, device work, clinical-supply commitments, and termination clauses can persist after a program changes. Internal personnel and stock compensation form a growing fixed base. Public filings do not provide a complete program-level schedule of cancellable and noncancellable obligations, so low property and equipment can understate economic capital intensity.

Earnings quality and cash conversion

Net loss and operating cash flow diverge because stock compensation and other noncash items are added back, while interest income reduces accounting loss. In H1 2026, stock compensation was $22.5 million—about 29% of operating cash burn and 21% of operating expenses. The cash conversion therefore looks better than net loss partly because employees receive equity. That is not free funding; it transfers future value and expands the economic denominator. [S1]

Interest income is economically valuable. More than $1 billion invested in Treasury, agency, and corporate securities partly offsets burn. The benefit is temporary: interest income falls as cash is consumed and may decline with rates. It should not be capitalized as a durable operating margin.

Collaboration cash can sometimes make a development platform appear self-funding. That transferable hypothesis is not load bearing here. Deferred revenue and cumulative license revenue are small relative to spending, while the cash balance arose principally from equity offerings. The accounting is not materially obscuring a profitable platform model because no such model exists.

Balance sheet and runway

At June 30, DNTH held $110.5 million of cash, $776.3 million of short-term investments, and $303.6 million of long-term investments, totaling $1.190 billion. Total liabilities were $41.2 million, with no conventional financial debt and approximately $1.3 million of lease liabilities. Management says available capital should fund operations into 2030. [S1]

Mechanical runway is longer than management’s planning statement: the June balance covers about 7.6 years at annualized H1 burn or 6.1 years at the Q2 exit rate. Management’s shorter horizon is more credible because late-stage trials, manufacturing, device, regulatory work, and commercial preparation should increase spending. The June balance also predates completion of the ATM; approximately 1.14 million additional common shares were outstanding by August 3, and the $200 million gross program had been exhausted. Subsequent proceeds must be offset against post-quarter burn and fees rather than simply added to June cash. [S1]

The absence of leverage sharply reduces insolvency and covenant risk. DNTH should be able to reach all announced near-term readouts without relying on a favorable financing market. It may still require capital before sustainable commercial self-funding, particularly if several programs succeed and require simultaneous launch investment. Large cash changes the timing of dilution risk, not its existence.

ROIC and research productivity

Reported ROIC is deeply negative because operating profit is negative. A secondary calculation places trailing ROIC near negative 18%, but the exact value is convention dependent because excess cash dominates the invested-capital denominator. Conventional ROIC is therefore descriptive rather than decision-useful at this stage. [S18]

The retrieved research-capital learning that accounting ROIC should be supplemented after a biotechnology company crosses into profitability is valid in its original scope but stale for DNTH. DNTH has not crossed into product revenue or profit. Capitalizing R&D now would convert failed and unvalidated research spending into an accounting asset and could create a misleading appearance of capital formation.

The appropriate sector-equivalent scorecard is cumulative cash consumed, diluted capital raised, milestone quality, remaining cash, and change in probability-adjusted asset value per share. On that scorecard, management produced randomized gMG evidence and parallel late-stage programs with cumulative operating burn below $400 million. The market now values the operating pipeline several billion dollars above cash, so the future burden of proof is much higher than it was during early development.

Capital intensity and contingent claims

Physical capital intensity is low: H1 capital expenditure was approximately $0.2 million. Economic capital intensity is high because trials, manufacturing, regulation, and failed-candidate risk are funded years before revenue and expensed immediately. DNTH212’s potential $962 million of milestones and royalties are success contingent but reduce economic returns. CRO, CDMO, device, cell-line, and commercial obligations create additional claims not represented by property and equipment. [S1][S2]

There is no meaningful working-capital cycle yet. Inventory, receivables, rebates, and distribution will emerge only near commercialization. Accordingly, current free cash flow is effectively operating cash burn less negligible capital expenditure; it is not a normalized conversion margin.

Verdict: Financial quality is strong in liquidity and weak in realized returns. The debt-free balance sheet protects the clinical plan, and reported cost growth is consistent with program advancement. Against that, burn is accelerating, stock compensation is material, contingent program economics are substantial, and no research capital has produced commercial cash. Runway is a strategic advantage—not an intrinsic-value floor at the current market capitalization.

Capital Allocation

Management’s defining capital-allocation achievement is serial equity financing at progressively higher prices. The defining cost is the multiplication of the economic share base.

Financing record and dilution

At the 2023 reverse-merger closing, private Dianthus combined its programs and financing with Magenta’s residual cash. Economic-basic shares at year-end 2023 were approximately 15 million after including pre-funded warrants. [S26][S30][S36]

The January 2024 financing sold approximately 14.5 million common shares and 4.7 million pre-funded warrants at a $12 common-equivalent price for $230 million gross. The September 2025 transaction sold approximately 7.63 million common shares and 1.11 million pre-funded warrants at a $33 equivalent price for $288.4 million. The March 2026 offering sold approximately 8.47 million common shares and 0.40 million pre-funded warrants at an $81 equivalent price for $719 million. ATM issuance added further shares and exhausted a $200 million program by August 2026. [S1][S27][S28][S29]

By August 2026, 55.934 million common shares and 1.347 million pre-funded warrants produced approximately 57.28 million economic-basic shares—about 3.8 times the year-end 2023 denominator. Calling this dilution without evaluating price and reinvestment would be incomplete. Raising $719 million at $81 is substantially better than funding the same work at $12. The correct test is whether each financing funds risk-adjusted value creation exceeding the value transferred to new investors.

Pre-funded warrants should be treated as economic common stock because their remaining exercise price is nominal. At June 30, 7.178 million options were outstanding at a $32.56 weighted strike, with 2.371 million exercisable. Approximately 1.895 million awards remained available for grant, and the plan’s annual evergreen had added another 2.16 million shares in January 2026. Unrecognized stock compensation was approximately $133.5 million over about three years. [S1]

Treasury-stock accounting captures current in-the-money options but not future evergreen grants, replacement awards, or another financing. A credible per-share valuation should therefore distinguish reported common shares, economic-basic shares, current treasury-method dilution, and expected future dilution.

Reinvestment and portfolio construction

Most cash has been reinvested in claseprubart trials, manufacturing, device development, personnel, and the DNTH212 license. This is internally coherent because the lead asset has the strongest evidence and can reuse infrastructure across indications. It also concentrates spending in one molecule before pivotal confirmation.

DNTH212 diversifies target biology and disease exposure but imports partner dependency, milestone leakage, royalties, and Greater China execution. DNTH312 is internally generated and may reuse platform knowledge, although moving it rapidly into the clinic before claseprubart pivotal resolution could increase both cash burn and correlated immune-mechanism risk.

DNTH has made no conventional post-merger acquisition, carries no acquisition debt, pays no dividend, and has not repurchased shares. A dividend would be inappropriate for a loss-making developer, and a buyback would conflict with the need to fund trials. DNTH is a Delaware corporation with ordinary common equity, not an ADR, partnership, MLP, or K-1 security. [S1][S2]

Governance and incentives

The 2026 proxy describes an independent board chair and an otherwise independent board except for CEO Marino Garcia. Countervailing provisions include a classified board, staggered terms, plurality voting, no cumulative voting, and change-in-control acceleration. The clawback policy is principally tied to accounting-restatement requirements. [S3]

For 2025, corporate goals weighted claseprubart at 45%, device and lifecycle work at 15%, pipeline expansion at 20%, and organization and funding at 20%. The compensation committee scored performance at 140% and added ten discretionary percentage points, producing a 150% bonus factor. Funding is a legitimate operating objective, but a gross capital-raised metric can reward denominator expansion unless paired with clinical value per share. [S3]

The CEO’s 2025 compensation included approximately $0.63 million salary, $0.52 million cash incentive, and $9.45 million of option grant-date value, for total disclosed compensation around $10.62 million. The options vest primarily with service rather than explicit pivotal, regulatory, or per-share-return hurdles. Options align executives with upside but can also reward volatility and market-cap growth despite dilution.

Insider activity

A review of recent Forms 4 through August 2026 identified grants, option exercises, tax withholding, and sales, but no code-P executive open-market purchase. The CEO exercised and sold shares in March under a previously adopted plan and exercised and sold another block in August; transaction-plan footnotes applied to some, but not necessarily every, August lot. Other executives also exercised and sold options. [S32][S33]

These transactions are not equivalent to discretionary bearish sales. Expiration, taxes, diversification, and 10b5-1 plans matter. Still, executives often sold rather than retained the newly acquired shares, and no open-market buying was identified. That is weaker alignment evidence than option-based beneficial-ownership percentages alone suggest.

Verdict: Capital allocation has been tactically strong: financing occurred after evidence and price improved, debt was avoided, and cash was directed toward the core portfolio rather than unrelated M&A. The disconfirming arithmetic is substantial dilution, rapidly rising stock compensation, and incentives that explicitly reward funding but not clearly rNPV per share. Future evaluation should emphasize risk-adjusted value created per diluted share, not cash raised or programs initiated.

Changes and Headwinds — Last Two Years

The last two years transformed DNTH from an early clinical developer into a well-funded late-stage company. The transformation improved both the opportunity and the quality of the risks.

In 2024, the $230 million financing supported parallel development and device work. R&D more than doubled as MaGic, MoMeNtum, and CAPTIVATE advanced, and internal personnel expanded. The strategy moved from single-program preservation toward multi-indication execution. [S25][S27]

In September 2025, MaGic produced the first randomized efficacy evidence. Both doses improved MG-ADL, QMG, and MGC relative to placebo. The same dataset disclosed dose-associated high-titer ANA and non-monotonic results on some secondary endpoints. Management’s best-in-disease framing became more plausible but remained unproven because no head-to-head trial existed. DNTH immediately raised $288.4 million at a higher price. [S4][S5][S28]

Late in 2025, DNTH licensed DNTH212 and began a partner-led China Phase 1 program, expanding beyond a single mechanistic platform. The transaction added diversification but also $30 million of paid cost, a possible $8 million near-term obligation, large success-based milestones, royalties, and partner dependency. [S1][S2]

In March 2026, CAPTIVATE reached its internal continuation threshold earlier than the maximum planned Part A enrollment. FDA also agreed to reduce ANA-related screening and monitoring in ongoing trials. The $719 million offering materially lowered financing risk and allowed parallel investment into late-stage trials and pipeline expansion. [S1][S5][S29]

In June, DNTH said EMERGE dosing began and presented detailed CAPTIVATE results. Two days before that presentation, Sanofi discontinued refractory-CIDP Phase 3 MOBILIZE. The temporal proximity is important: DNTH’s own early evidence improved while the closest class precedent worsened. The rational posterior is neither full validation nor class abandonment, but a higher evidentiary hurdle for CAPTIVATE Part B. [S5][S6]

In August, DNTH reported $1.190 billion of June liquidity, reiterated runway into 2030, confirmed December MoMeNtum guidance, and introduced DNTH312. H1 stock compensation of $22.5 million nearly equaled the full-year 2025 amount, while both R&D and G&A continued increasing. [S1][S16]

Management commentary is available principally through SEC-filed releases, presentations, and prepared conference materials. Company Financials returned no DNTH earnings-call transcripts for the latest two reported quarters, and no substantive public quarterly Q&A was identified. That absence does not imply concealment, but it limits independent review of questions concerning CAPTIVATE’s patient funnel, refractory subgroup, ANA persistence, Q4W support, and dilution.

The operating environment also changed. The 2025 10-K describes U.S. biotechnology-supply-chain restrictions and exposure to foreign CRO/CDMO relationships, including providers that may be affected by policy. A required supplier transfer could introduce comparability, validation, capacity, and timing risk. [S2]

Other headwinds include crowded gMG launches, evolving CIDP standards, competing MMN data, global enrollment, manufacturing scale-up, patent prosecution, and commercial-preparation expense. There is no disclosed financial-leverage crisis or material litigation dominating the thesis. The principal recent deterioration is higher embedded expectations alongside adverse same-target CIDP evidence.

Verdict: Execution over the last two years has been strong in financing, trial initiation, and portfolio expansion. Evidence quality also became more complicated: ANA emerged, Q4W remains unrandomized, same-target CIDP evidence failed at Phase 3, and spending plus dilution accelerated. Prepared management confidence should be treated as a hypothesis to test against controlled outcomes.

Risk Analysis

DNTH’s risk is dominated by clinical discontinuity rather than economic cyclicality or leverage. License revenue and R&D timing can make quarterly losses uneven, but ordinary GDP cycles do not determine intrinsic value. Rates and biotechnology risk appetite influence the stock and interest income; clinical results, safety, labels, patents, and competition determine the long-run outcome.

Catastrophic and total-loss paths

A literal zero is unlikely while DNTH holds substantial cash and multiple programs. Severe permanent impairment is plausible. The clearest path is a molecule-level safety issue affecting all claseprubart indications, followed by failure or abandonment of the early pipeline and several years of shutdown or restructuring costs. A combined EMERGE and CAPTIVATE efficacy failure without a safety problem would still eliminate most operating value and could move the enterprise toward residual net cash.

The predecessor’s history illustrates the speed of such a reset but not its probability for current DNTH. Magenta’s safety failure involved a different asset and organization. It is relevant as capital-market history, not biological evidence. [S31]

A subtler risk is successful development without sufficient differentiation. Claseprubart could gain approval with Q2W-only dosing, similar efficacy to incumbents, vaccination or monitoring requirements, and large rebates. That outcome may produce revenue while failing to justify current enterprise value. Valuation risk therefore extends well beyond outright trial failure.

Cyclicality, rates, and the factor model

There is no stable earnings cycle to forecast. Investment income is rate sensitive, and trial payments create quarter-to-quarter volatility. Risk appetite can affect financing terms, although DNTH’s current cash makes near-term issuance less necessary.

The factor model dated September 2 reports exposures of -0.977 to liquidity, 0.639 to market, 0.571 to health care, 0.531 to growth, -0.510 to interest rates, and 0.373 to small size. Sector values are statistical return exposures, not industry classifications. The liquidity and interest-rate signs cannot be given a causal interpretation without model-convention detail. [S24]

The model’s R² is only 0.114 and adjusted R² 0.098, implying that roughly 89% of historical return variation was not explained by included factors. Residual volatility is 0.714 and residual Sharpe 0.992, but frequency and annualization conventions are unavailable. The decision-useful conclusion is limited: clinical events dominate, and factor-adjusted historical performance should not be extrapolated as evidence of asset quality.

Verdict: Balance-sheet insolvency risk is low, while clinical, safety, competitive, and valuation risks remain high. The most asymmetric downside is molecule-level impairment across three indications. The more probable but less dramatic danger is approval with insufficient dosing, label, price, or efficacy differentiation to earn the economics embedded in the stock.

Valuation Discussion

Conventional multiples provide little information. License revenue is immaterial, earnings and free cash flow are negative, and book value is dominated by recently issued cash. A revenue multiple would capitalize collaboration accounting rather than product demand. A book multiple measures market expectations for intellectual property, not operating profitability.

Capitalization and reconciliation

At $107.85 and 55.934 million reported common shares, common market capitalization was approximately $6.03 billion. Adding 1.347 million pre-funded warrants gives $6.18 billion of economic-basic equity value. Subtracting $1.190 billion of June cash and investments gives approximately $4.99 billion of economic-basic enterprise value. [S1][S18]

A secondary market source reports conventional enterprise value around $4.84 billion because it uses reported common market capitalization and subtracts the cash and investment balance without adding the pre-funded-warrant economics. The difference is definitional rather than an operating disagreement. [S18]

Applying the treasury-stock method to 7.178 million options at a $32.56 weighted strike adds approximately five million net shares at the current price. That produces roughly 62.3 million current diluted economic shares and diluted enterprise value near $5.5 billion. This still omits future employee awards, evergreen additions, and any later financing. The June cash balance also predates post-quarter ATM proceeds and subsequent burn.

Cash equals approximately $20.8 per economic-basic share or about $19 per current treasury-method diluted share. Cash therefore provides operational endurance but protects only a small fraction of the quoted price.

Peer context

Peer selection should reflect risk rather than superficial biotechnology classification. Annexon is a complement-focused clinical developer, Zenas is a clinical immunology developer, and Immunovant is a larger late-stage autoimmune company. Argenx is not a development-stage valuation peer; it is evidence of what successful gMG/CIDP commercialization can become.

Recent secondary-market estimates place enterprise value around $0.73 billion for Annexon, $1.69 billion for Zenas, and $7.36 billion for Immunovant. DNTH’s economic-basic EV is almost three times the median of those development-stage references, while remaining below Immunovant. Different assets, cash dates, indications, ownership structures, and development stages make this a context range, not a direct multiple. [S19][S20][S21]

Argenx’s enterprise value exceeds $59 billion, with trailing revenue above $5 billion and positive operating cash generation. Its approximate 11-times EV/sales multiple validates the potential value of successful autoimmune commercialization but cannot be applied to unapproved DNTH sales. Argenx has established manufacturing, reimbursement, a safety database, and commercial execution that DNTH has not yet built. [S22]

The appropriate interpretation is that DNTH deserves a premium to many early developers because it has controlled gMG activity, multiple late-stage indications, and more than $1 billion of liquidity. The premium also means ordinary Phase 3 attrition or a merely adequate label can destroy more value than at a cash-near peer.

Scenario rNPV

The following is an analyst estimate, not management guidance. It models indication-level peak sales, success probability, launch timing, post-commercial contribution margin, patent life, future spending, and dilution. Sales ramp from roughly 10% of peak toward maturity over seven years. Bear, base, and bull discount rates are 13%, 11.5%, and 10%; contribution margins are 50%, 55%, and 60%. Claseprubart economics are modeled through approximately 2042–2044, consistent with—but not guaranteed by—the company’s stated patent expectations. [S2]

CIDP probability is deliberately below the draft’s original 55% central assumption because MOBILIZE is relevant negative evidence. It remains well above zero because CAPTIVATE addresses a broader population and uses a different molecule and design. DNTH212 economics are reduced for milestones and royalties rather than treating gross sales as wholly owned. [S1][S6]

Future development and corporate spending is deducted at approximately $1.2 billion in the bear case, $0.9 billion in the base case, and $0.8 billion in the bull case on a present-value basis. Lower bull present-value spending reflects earlier success and cash-flow arrival, not an assumption that commercialization is inexpensive. Future diluted shares are modeled at 75 million in the bear case and 70 million in base and bull cases.

The values are sensitive enough that the assumptions matter more than the last dollar. Using current treasury-method shares instead of future dilution would raise the same central equity value into the mid-$90s per share. Lowering CIDP probability, delaying launch one year, or adding commercial expense can erase several hundred million dollars. Successful Q4W dosing and clean multi-indication safety can add several billion.

What the current price embeds

Current diluted enterprise value is approximately equal to the modeled central product rNPV before fully deducting future corporate spending and dilution. Supporting the present quotation therefore requires some combination of aggregate claseprubart peak sales above $4 billion, success in at least two meaningful indications, launch around 2029–2030, strong contribution margins, a favorable chronic-use label, patent durability into the early 2040s, and some early-pipeline value.

The market correctly recognizes that near-term financing risk is unusually low and that MaGic is more than mechanistic speculation. It may also be correct that infrequent self-administration creates substantial commercial value. The fragile assumptions are CIDP probability after MOBILIZE, Q4W efficacy, long-duration safety, payer capture of convenience value, and future dilution.

The bear outcome need not be complete failure. One successful indication with $1–$2 billion peak sales, Q2W-only dosing, ordinary rebates, a 2031 launch, or a monitoring-heavy label could all generate substantial revenue while producing value below the present capitalization. Conversely, clean Q4W approval in both gMG and CIDP could support value materially above the central scenario.

Own-history context

There is no meaningful long-duration own-history multiple because the current operating company began trading through the 2023 reverse merger and remains precommercial. The useful history is event-adjusted enterprise value: the market moved from valuing an early lead asset near cash to valuing a multi-indication franchise billions above cash. That history reflects changing probabilities, not multiple expansion on recurring fundamentals.

Verdict: Diluted rNPV is the appropriate framework, and its width is the investment fact. The balance sheet funds the evidence-generation plan, but cash represents only about one-fifth of the share price. At current capitalization, investors are principally underwriting successful pivotal development and commercial differentiation rather than purchasing cash-protected optionality.

Variant Perception

Apparent consensus

The market’s apparent view is that claseprubart combines competitive efficacy, infrequent self-administration, and a potentially cleaner complement-safety profile; that CAPTIVATE’s 75% response materially de-risks CIDP; and that more than $1 billion of liquidity removes financing as a central concern. The share-price appreciation and development-peer premium are consistent with that interpretation, although no formal sell-side consensus dataset is treated as primary evidence. [S5][S18]

Strongest bull case

The bull case begins with concordant MaGic activity across MG-ADL, QMG, and MGC, rapid symptom improvement, and a more favorable laboratory-safety profile at 300 mg than 600 mg. Claseprubart could deliver competitive disease control with Q2W or Q4W convenience and less burdensome infection management than C5 inhibition. CAPTIVATE adds evidence of CIDP biological activity, while MMN can create a third indication. Shared manufacturing, device, safety, and neurologist infrastructure create franchise leverage. DNTH212 and DNTH312 then expand the platform beyond neuromuscular disease. [S1][S4][S5]

The balance sheet makes this case more credible because management need not ration every program around one financing window. Clinical setbacks can be investigated rather than immediately forcing a strategic transaction. Serial offerings at $12, $33, and $81 demonstrate that management captured appreciation to fund more milestones with progressively fewer shares per dollar. [S27][S28][S29]

Strongest bear case

The bear case is that the market has mistaken promising exploratory and open-label evidence for pivotal validation. MaGic was small, Q4W was not randomized, and several secondary results lacked clean dose response. High-titer ANA was dose associated, and extended half-life makes unexpected toxicity less reversible. [S4][S5]

CAPTIVATE transitioned patients rapidly from immunoglobulin, used a one-point adjusted-INCAT response threshold, selected responders into randomized withdrawal, and has not disclosed outcomes by prior-treatment stratum or the full screening funnel. Riliprubart’s sequence demonstrates that favorable open-label active-C1s evidence can fail in controlled refractory CIDP. [S5][S6][S7]

gMG is crowded, CIDP is becoming more competitive, and commercial success requires superiority on some combination of administration, safety, efficacy, persistence, or net economics. A roughly $5.5 billion diluted enterprise value leaves limited room for a single ordinary approval.

Load-bearing assumptions

1. The 300-mg regimen reproduces MaGic. This assumption fails if EMERGE produces a materially smaller MG-ADL effect, objective QMG does not corroborate it, discontinuation is high, or placebo response overwhelms separation. [S4][S12]

2. Q4W maintains commercially adequate efficacy. This fails if trough exposure, rescue, or clinical outcomes are meaningfully inferior to Q2W. [S5][S12]

3. CAPTIVATE Part A reflects drug activity that survives randomization. This fails if continued claseprubart and placebo have similar relapse hazards, refractory patients perform poorly, or only a narrow fraction of screened patients reaches randomization. [S5][S6][S13]

4. Pathway selectivity produces a cleaner commercial label. This fails with clinical drug-induced lupus, disproportionate serious infection, mandatory monitoring, or vaccination and risk-management requirements that erase the convenience advantage. [S5][S8][S9]

5. Dilution remains subordinate to asset-value growth. This fails if the economic share count, employee equity, and cash burn grow faster than probability-adjusted pipeline value. [S1][S3]

Positioning and factor context

The factor model shows material health-care, growth, market, liquidity, interest-rate, and small-size exposures, but explains only 11.4% of return variance. Sector betas are statistical exposures, not legal or operating classifications, and the unknown sign conventions limit causal interpretation. Clinical catalysts and residual volatility dominate position risk. [S24]

Revalidated, contradicted, and stale learnings

The retrieved learning that a small efficacy trial cannot establish safety differentiation against a rare labeled risk is directly applicable. MaGic and the healthy-volunteer study can identify common tolerability and pharmacology but cannot exclude rare infection or autoimmune events. FDA’s reduced ANA monitoring is favorable process evidence, not a statistical substitute for exposure.

The research-adjusted-ROIC learning for biotechnology companies transitioning into profitability is valid generally but outside DNTH’s current scope. Applying it now would capitalize unvalidated R&D and overstate asset certainty.

The hypothesis that collaboration upfront cash can distort operating cash flow is not material here. Collaboration revenue is small; equity issuance is the source of balance-sheet strength.

Unrelated retrieved learnings concerning utilities, acquisitions, streaming, telecommunications, retail fuel, and medical-device infrastructure have no evidentiary connection to DNTH and are rejected rather than transferred by analogy.

The most useful company-specific variant is not that the clinical data are poor. They are promising. It is that the present market value may assign pivotal-quality probability to evidence that remains small, uncontrolled, modeled, or challenged by same-target failure.

Verdict: Bulls have the stronger molecule narrative; bears have the stronger evidence-quality and valuation argument. The rational stance is to respect MaGic while applying a larger discount to CIDP, Q4W, and rare-event safety than a headline reading of CAPTIVATE implies.

Fact vs. Interpretation

Verdict: Cash, trial design, measured changes, ANA incidence, financing, and the competitor discontinuation are reported facts. Best-in-disease efficacy, a light commercial label, broad CIDP benefit, Q4W equivalence, and multi-billion-dollar penetration are hypotheses. The valuation depends principally on the latter group becoming facts.

Open Questions

1. What proportion of screened CAPTIVATE patients entered Part A, met response criteria, entered Part B, and remained evaluable? A responder percentage without the complete funnel can overstate commercial eligibility. [S5][S13]

2. What were CAPTIVATE responses separately for standard-of-care-treated, refractory, and treatment-naïve patients? The answer is particularly important after MOBILIZE’s refractory result. [S5][S6]

3. What is CAPTIVATE Part B’s event target, power assumption, censoring plan, and expected readout window? Public disclosure is insufficient for an independent event-accumulation model.

4. How much immunoglobulin activity remains after the seven-day transition, and how might it influence early disease trajectory, rescue decisions, and adjusted-INCAT interpretation?

5. How robust is 300-mg Q4W exposure by body weight, disease activity, geography, and immunogenicity? Modeled average exposure can conceal trough-risk tails. [S5][S12]

6. What is EMERGE’s multiplicity hierarchy across Q2W, Q4W, MG-ADL, QMG, response, and minimal symptom expression?

7. Will high-titer ANA persist with multi-year exposure, and will adjudicated clinical drug-induced lupus remain absent? Reduced routine testing increases the importance of standardized long-term sampling and event adjudication. [S1][S5]

8. What vaccination, infection-monitoring, warning, or risk-management profile does FDA consider likely for chronic active-C1s inhibition? Development aspirations are not a negotiated commercial label. [S8][S9]

9. Can MoMeNtum show consistent benefit in rescue timing, grip strength, MRC score, and MMN-RODS despite its small, safety-primary design? [S14]

10. What exact DNTH212 dataset will be disclosed in 2026—single-dose safety, PK, BDCA2/pDC engagement, interferon pharmacology, BAFF/APRIL effects, immunoglobulin changes, or a subset? [S15]

11. How much manufacturing, device, and commercial-readiness spending is included in runway into 2030, and which programs would be delayed if cost exceeds plan?

12. Which issued claims directly cover claseprubart composition and active-state binding, and how broad are method, formulation, device, and foreign-jurisdiction protections? [S2]

13. Will future executive equity include pivotal, regulatory, or per-share-value conditions rather than relying mainly on service-vesting options and funding goals? [S3]

14. What are executives’ unoptioned common holdings after recent exercise-and-sale transactions, and what explains the limited retention of exercised shares? [S33]

15. Does DNTH plan to commercialize alone in the United States, partner ex-U.S., or license before approval? Each path changes retained margin, capital need, and execution risk.

16. Will future quarterly updates include an archived Q&A or transcript addressing trial design, safety, and capital allocation? Prepared statements alone provide limited adversarial scrutiny.

Verdict: None of these questions independently invalidates the thesis. Together they show where apparent precision exceeds public evidence—particularly CAPTIVATE’s funnel and subgroup activity, Q4W support, long-duration ANA outcomes, patent breadth, and the spending required to convert a funded pipeline into a commercial franchise.

What Must Be True

Bull tests

1. MaGic must replicate. EMERGE’s 300-mg Q2W regimen should produce a clinically meaningful, statistically robust MG-ADL effect with objective QMG corroboration. Q4W should remain sufficiently close to Q2W to support a genuine convenience advantage. [S4][S12]

2. CIDP benefit must survive randomization. CAPTIVATE Part B should show a convincing time-to-relapse advantage, acceptable absolute relapse rates, and a commercially relevant screening-to-randomization yield. Benefit should not be confined to a narrow nonrefractory subgroup. [S5][S6][S13]

3. Safety differentiation must become exposure based. Larger and longer datasets should continue to show no clinical DIL/SLE, no disproportionate serious bacterial infection, and no monitoring, vaccination, boxed-warning, or REMS burden that erases administration convenience. [S5][S8][S9]

4. MMN must add independent value. MoMeNtum should delay immunoglobulin rescue and show concordant grip, MRC, and MMN-RODS benefit. A safety-only or post hoc narrative is insufficient. [S1][S14]

5. DNTH212 must validate both mechanisms at a usable dose. The same dose should demonstrate pDC/type-I-interferon and BAFF/APRIL pharmacology without excessive immunoglobulin reduction, infection, cytokine reaction, or immunogenicity. [S1][S15]

6. Commercial differentiation must survive competition. Product labels, payer access, persistence, and physician adoption should show that infrequent active-C1s inhibition is more than another similarly effective premium biologic. [S8][S10][S11]

7. Capital allocation must create value per share. Fully diluted rNPV should rise faster than cash use, evergreen additions, options, and equity issuance. [S1][S3]

Bear tests

1. The gMG signal proves fragile. The bear case strengthens if EMERGE’s effect is materially smaller than MaGic, QMG fails to corroborate MG-ADL, placebo response remains high, or Q4W is inferior. [S4][S5]

2. CAPTIVATE follows the riliprubart pattern. The bear case strengthens if open-label responders relapse similarly on continued drug and placebo, refractory patients show little activity, or the randomized cohort represents only a narrow fraction of screened patients. [S5][S6][S7]

3. Safety compromises convenience. Persistent ANA, clinical DIL, serious infection, prolonged adverse events, vaccination requirements, or ongoing laboratory monitoring would erode the proposed moat. The long half-life would then increase rather than reduce economic risk. [S2][S5][S9]

4. MoMeNtum is equivocal. Similar rescue timing to placebo, inconsistent objective measures, or clearly superior competitor evidence would reduce MMN option value. [S14][S34][S35]

5. Early-pipeline optionality fails. DNTH212 loses value if both intended pharmacodynamic effects cannot be obtained at a safe chronic dose. DNTH312 should remain near zero if it does not enter Phase 1 on schedule or offers no human differentiation. [S15][S16]

6. Economics disappoint despite approval. The bear thesis can be correct with regulatory success if launch slips beyond 2031, peak sales remain near $1–$2 billion per major indication, gross-to-net deductions are heavy, commercial expense is high, or patent protection narrows. [S1][S2]

7. Dilution absorbs progress. A continued rise toward or above 75 million diluted shares without proportional rNPV growth would reduce per-share outcomes even while solvency remains strong. [S1][S3]

Monitoring dashboard

The highest-value quarterly signals are CAPTIVATE’s patient funnel, randomization and subgroup response; EMERGE site activation and enrollment; MoMeNtum rescue timing plus objective motor measures; ANA persistence and adjudicated clinical events; serious infections and discontinuations; Q4W pharmacokinetics; DNTH212 dual pharmacodynamics; operating cash burn net of interest timing; economic diluted shares; unrecognized stock compensation; and competitor results in MMN, gMG, CIDP, SLE, and Sjögren disease. [S1][S5][S12][S13][S14][S15]

The clinical thesis is falsified on the downside by a molecule-level safety issue that compromises chronic use or by controlled efficacy failure in both CAPTIVATE and EMERGE. The cautious valuation thesis is falsified on the upside if randomized CIDP succeeds convincingly, Q4W reproduces gMG efficacy, long-duration safety remains clean, and diluted rNPV rises substantially without equivalent share issuance.

Claseprubart does not need to dominate every endpoint or indication. The present valuation does require more than one ordinary success: it must become a durable and commercially differentiated franchise across at least two meaningful indications while management converts abundant cash into value per diluted share.

Core evidence links: S1 — Q2 2026 filing, S2 — 2025 annual filing, S4 — MaGic results, S5 — CAPTIVATE and MaGic presentation, and S6 — MOBILIZE discontinuation.

Public source appendix

• S1: Dianthus Therapeutics Q2 2026 Form 10-Q — SEC Form 10-Q; published 2026-08-04; Financial statements; equity and stock compensation notes; license obligations; clinical updates; liquidity and MD&A
• S2: Dianthus Therapeutics 2025 Form 10-K — SEC Form 10-K; published 2026-03-09; Business, mechanism, patents, employees, competition, suppliers, risk factors, and audited financial statements
• S3: Dianthus Therapeutics 2026 Proxy Statement — SEC definitive proxy; published 2026-04-09; Governance; beneficial ownership; incentive-goal weights and scoring; executive-compensation tables
• S4: Dianthus Phase 2 MaGic Top-Line Results — SEC-filed clinical release; published 2025-09-08; Trial population, efficacy tables, responder measures, safety, and one-sided statistical notes
• S5: CAPTIVATE and MaGic Clinical Presentation — SEC-filed company presentation; published 2026-06-12; MaGic two-sided analyses and ANA; Q4W rationale; CAPTIVATE design, baseline strata, outcomes, and competitive assays
• S6: Sanofi Discontinues Phase 3 MOBILIZE Study — Competitor primary release; published 2026-06-10; Independent monitoring committee efficacy conclusion, refractory-CIDP scope, and absence of a new safety signal
• S7: Sanofi Riliprubart Phase 2 CIDP Results — Competitor primary release; published 2024-06-25; Open-label efficacy and safety results by standard-of-care, refractory, and treatment-naïve cohort
• S8: FDA Zilucoplan Prescribing Information — FDA prescribing information; publication date unavailable; Boxed warning, meningococcal infection, vaccination, REMS, and administration
• S9: FDA Enjaymo Prescribing Information — FDA prescribing information; publication date unavailable; Serious-infection warning, vaccination recommendations, and autoimmune-disease discussion
• S10: FDA Drug Trials Snapshot: Imaavy — FDA review summary; publication date unavailable; gMG trial population, design, MG-ADL effect, and QMG effect
• S11: FDA Approval Summary for Vyvgart Hytrulo in CIDP — FDA approval summary; published 2024-06-21; Open-label responder identification, randomized withdrawal, efficacy conclusion, and weekly administration
• S12: EMERGE Phase 3 gMG Trial — NCT07647510 — ClinicalTrials.gov registry; publication date unavailable; Enrollment, Q2W/Q4W/placebo arms, endpoints, eligibility, and estimated dates
• S13: CAPTIVATE Phase 3 CIDP Trial — NCT06858579 — ClinicalTrials.gov registry; publication date unavailable; Open-label Part A, responder criteria, randomized-withdrawal Part B, endpoints, and estimated dates
• S14: MoMeNtum Phase 2 MMN Trial — NCT06537999 — ClinicalTrials.gov registry; publication date unavailable; Arms, enrollment, safety objective, rescue criteria, motor and functional measures, and dates
• S15: DNTH212/LBL-047 Phase 1 Trial — NCT07323173 — ClinicalTrials.gov registry; publication date unavailable; Healthy-volunteer and SLE portions, endpoints, enrollment, and estimated completion
• S16: Dianthus Announces DNTH312 — Company release; published 2026-08-04; Mechanism, preclinical observations, cross-study comparator limitations, and Phase 1-readiness target
• S17: Dianthus August 2026 Corporate Presentation — Company presentation; published 2026-08-01; Pipeline, market estimates, development timelines, and capitalization
• S18: DNTH Statistics, Valuation, and Market Data — Secondary market data; published 2026-09-03; September 3, 2026 close; market capitalization; enterprise value; 52-week context; ratios and returns
• S19: Annexon Statistics and Valuation — Secondary peer market data; published 2026-09-03; September 2026 enterprise value, capitalization, cash, and clinical-stage financial context
• S20: Zenas BioPharma Financial Ratios and Valuation — Secondary peer market data; publication date unavailable; August 2026 enterprise value and capitalization
• S21: Immunovant Financial Ratios and Valuation — Secondary peer market data; publication date unavailable; September 2026 enterprise value and capitalization
• S22: argenx Statistics and Valuation — Secondary commercial-peer market data; publication date unavailable; September 2026 enterprise value, trailing revenue, profitability, cash flow, and valuation multiples
• S23: Dianthus Therapeutics 2024 Form 10-K — SEC Form 10-K; published 2025-03-11; Audited financial statements, R&D history, equity issuance, and stock compensation
• S24: factor model — DNTH snapshot — Quantitative factor diagnostic; published 2026-09-02; Exposures, residual signals, R², adjusted R², and alpha dated September 2, 2026
• S25: Dianthus Therapeutics 2023 Form 10-K — SEC Form 10-K; published 2024-03-21; Reverse recapitalization, post-merger capitalization, business history, and audited financial statements
• S26: Dianthus Therapeutics 2024 Form 10-K — SEC Form 10-K; published 2025-03-11; Year-end 2023 and 2024 capitalization, financing, operating history, and audited statements
• S27: Dianthus January 2024 Private Placement Form 8-K — SEC Form 8-K; published 2024-01-22; Common shares, pre-funded warrants, offering prices, and gross proceeds
• S28: Dianthus September 2025 Offering Form 8-K — SEC Form 8-K; published 2025-09-11; Common shares, pre-funded warrants, offering prices, and gross proceeds
• S29: Dianthus March 2026 Offering Prospectus — SEC Rule 424(b)(5) prospectus; published 2026-03-11; Offering summary, common shares, pre-funded warrants, capitalization, proceeds, and dilution
• S30: Magenta–Dianthus Reverse Merger Announcement — SEC Form 8-K; published 2023-05-03; Transaction structure, concurrent financing, expected cash, and operating-company change
• S31: Magenta MGTA-117 Safety Update — SEC-filed predecessor-company release; published 2023-01-25; Patient death, serious adverse events, enrollment pause, and investigation
• S32: DNTH SEC Filing Index — SEC filing index; publication date unavailable; Trailing filing inventory, including recent Forms 3, 4, and 5
• S33: Marino Garcia August 2026 Form 4 — SEC Form 4; published 2026-08-07; Option exercise, sales, post-transaction holdings, and trading-plan footnotes
• S34: argenx 2024 Annual Report — Empasiprubart Development — Competitor annual report; published 2025-03-01; ARDA Phase 2 MMN results and development strategy
• S35: EMPASSION Phase 3 MMN Trial — NCT06742190 — ClinicalTrials.gov registry; publication date unavailable; Head-to-head design, enrollment, endpoints, comparator, and estimated dates
• S36: Dianthus Predecessor Audited 2021–2022 Financial Statements — SEC-filed audited financial statements; published 2023-09-21; Balance sheets, statements of operations, cash flows, and preferred-financing notes