Coherus Oncology (CHRS) — Scientific Deep Dive for LOQTORZI, Casdozokitug, and Tagmokitug

Executive Summary

The Hook. Coherus divested every profitable biosimilar it had and bet the company on three immuno-oncology assets: a PD-1 inhibitor (LOQTORZI) in-licensed with a brutal low-20%-range royalty back to Junshi Biosciences, plus two mid-stage antibodies inherited from the 2023 Surface Oncology acquisition — casdozokitug (anti-IL-27) and tagmokitug (anti-CCR8 Treg depleter). The casdozo-atezo-bev triplet hit a 38% ORR and 17.2% CR in 29 first-line HCC patients at ASCO-GI 2025, numbers that look bio-better against IMbrave150’s ~30% ORR / ~7-8% CR, but it was open-label and uncontrolled. The randomized Phase 2 readout in mid-2026 is the binary event that determines whether CHRS is a pipeline-in-a-product or a dilution machine.

The Bull Case. If the mid-2026 casdozo/tori/bev triplet readout confirms a 30%+ CR rate over the tori/bev doublet, Coherus owns the only clinical-stage IL-27 antagonist with anti-tumor activity and a first-in-class profile in a $5–7B global 1L HCC opportunity. Layer tagmokitug — which hit a >50% intratumoral CCR8+ Treg depletion and 73% increase in CD8+ T cell density at SITC 2025, with a confirmed PR in a PD-1 refractory HNSCC patient at AACR 2025 — and you have a credible CCR8 program in an attractive M&A category where BMS, Gilead/Jounce, AbbVie, Roche, and Amgen are all developing competitors. Add the Janssen pasritamig combo deal for tagmokitug in mCRPC (Feb 2026 supply agreement), potential UDENYCA earnouts of $37.5M each (Sept 30, 2026 and Mar 31, 2027 sales hurdles), and an in-flight LOQTORZI franchise that grew 61% YoY to $11.8M in Q1 2026, and the SOTP can easily justify a $5–10 stock vs. ~$1.80 spot.

The Bear Case. The structural problem is that Coherus doesn’t own LOQTORZI — Junshi does. Coherus pays “a royalty in the low twenty percent range on net sales of LOQTORZI and up to an aggregate $380.0 million in one-time payments for the achievement of various regulatory and sales milestones”. Of that, ~$355M remained unpaid as of March 31, 2026. NPC is a small disease (~2,000 new US cases per year per GLOBOCAN 2022), the US NPC market is roughly $75–90M today growing to ~$120–130M by 2030, and Penpulimab (Akeso) was approved by FDA for NPC in April 2025 — meaning LOQTORZI’s “only approved” monopoly is over. Cash is $167M, Q1 burn from continuing operations was $36.9M, and management already raised $53.6M in a Feb 2026 follow-on at $1.75 — at a 49% discount to the Guggenheim $12 PT. Translation: management has demonstrated willingness to dilute pre-catalyst. Cross-trial comparisons of single-arm casdozo data to IMbrave150 historical controls have failed many oncology programs (LEAP-002 and COSMIC-312 both missed against the same atezo/bev benchmark). Tagmokitug’s “promising early activity” is one confirmed PR in 21 evaluable patients — that’s noise, not signal.

Bottom Line. This is a binary-event lottery ticket with structural royalty bleed. The mid-2026 casdozo Phase 2 readout could be one of the best risk/reward in the small-cap I/O space if you accept that ~50% of the float could be issued before commercial cash flow turns positive.

Catalyst Calendar & Financial Runway

Upcoming Catalysts (next 12–18 months)

• Mid-2026 — Casdozokitug 1L HCC Phase 2 randomized readout (CHS-388-202, NCT06679985). This is the binary event. The trial is randomized 1:1:1 across three arms (~72 patients): casdozo 700mg + tori + bev, casdozo 1400mg + tori + bev, and tori + bev alone (the internal control). Patient accrual is complete. Primary endpoints include ORR and Project Optimus dose-finding; the key read is whether casdozo arms separate from tori/bev control on CR rate and PFS. Stock reaction here will be ±50% based on similar single-asset I/O readouts.

• Mid-2026 — Tagmokitug Phase 1b dose optimization in 2L HNSCC and 2L upper GI adenocarcinoma. Initial efficacy data for the CHS-114 + toripalimab combination. Set expectations: this is a dose optimization study and the headline will be ORR/safety in small cohorts (n=40 HNSCC expansion), not a registrational endpoint. Anything ≥20% ORR in 2L HNSCC with manageable safety would be a positive signal.

• 2H 2026 — Tagmokitug Phase 1b in 1L/2L ESCC ± chemo.

• 2H 2026 — Tagmokitug Phase 1b/2a in 4L+ colorectal cancer (MSS). This is the longest reach — MSS CRC is checkpoint refractory and almost everything has failed here. A signal would be franchise-defining.

• Fall 2026 — Tagmokitug + pasritamig Phase 1b initiation in mCRPC under the Janssen clinical supply agreement signed Feb 4, 2026. This is a free-look optionality play — Janssen supplies pasritamig, Coherus sponsors the trial, both retain commercial rights.

• Sept 30, 2026 — UDENYCA Earnout #1 trigger date. $37.5M payable if UDENYCA net sales (under Intas/Accord) hit $300M for the four quarters ending Sept 30, 2026. Earnout #2 of $37.5M triggers if net sales hit $350M through March 31, 2027.

The Dilution Gap

Cash, cash equivalents, and marketable securities were $167.0 million at March 31, 2026. Quarterly metrics from continuing operations were a $36.9M loss and $57.9M cash used in operations — though the latter is inflated by $11.3M inventory build (commercial supply for the onshoring initiative), a $3.3M TSA wind-down payment, and other one-time working capital movements. A clean run-rate continuing-ops cash burn is closer to $25–35M per quarter.

At $30M/quarter, $167M provides 5–6 quarters of runway, putting the cash-out date in mid-2027 — which is after the casdozo binary readout but before any post-Phase-2 commercial scale-up could realistically begin. Management’s going-concern statement says cash “will be sufficient to fund our planned expenditures and meet our obligations for at least the twelve months following our financial statement issuance date” — i.e., through May 2027. That’s exactly long enough to clear the mid-2026 readout.

Importantly, Coherus has $64.9 million remaining on its TD Cowen ATM facility and just demonstrated tap appetite by issuing 32,890,000 shares at $1.75 on Feb 12, 2026 for $53.6M net proceeds. The Feb 2026 offering happened at the spot price — there’s no warehoused capital here, and a positive readout could be met with a follow-on at the post-pop level. A raise is more likely than not before commercial cash flow turns positive, and the only open question is whether it lands pre- or post-readout. Given the cash position and the binary nature of mid-2026, pre-readout raises become harder to execute; expect tap-the-ATM dribbles, not a marketed deal, until data clears.

Debt is modest: $38.7M 2029 Term Loan at SOFR + 8% (interest-only quarterly, matures May 2029) plus a $14.4M revenue participation liability with Coduet Royalty (5% of LOQTORZI US net sales, capped at 2.25x of the $37.5M LOQTORZI purchase price). The 2026 Convertible Notes were paid off in April 2026 ($121K residual). The 2029 Term Loan does carry a minimum cash covenant, which is a quiet killer — running the balance sheet too low triggers default before bankruptcy is mathematically required.

Insiders & Institutions

Holder composition is unusual for a small-cap biotech: there is no specialist-fund cluster. The 13G filers are individual/family trusts and passive index shops, not the Baker Bros/RA Capital/Perceptive/Vivo/Fairmount stack you’d expect on a high-conviction binary trade.

• Timothy G. Youngquist 2020 Irrevocable Trust: 12,790,800 shares = 8.27%. The filing notes 500,000 shares are call options.

• BlackRock: 7,337,261 shares = 4.9% as of March 31, 2026. Just under the 5% threshold.

• Thomas A. Satterfield, Jr. (aggregated with Tomsat Investment & Trading, Caldwell Mill Opportunity Fund, A.G. Family L.P.): 6,930,000 shares = 4.6%.

The absence of a biotech-specialist anchor (no Baker, no RA, no Perceptive, no Fairmount, no Avoro) is a yellow flag — these funds typically pile in ahead of high-conviction Phase 2 binaries. Either the casdozo readout is being viewed as a coin flip by smart money, or they’re waiting for a post-readout discounted secondary to build positions cheaply. Bears may call this the dog that didn’t bark. Total institutional ownership runs in the 70%+ range, so floats aren’t tightly held by retail — index funds and a long tail of small institutional holders dominate.

Coherus repurchased zero of its own equity during Q1 2026 (72,078 shares were withheld for tax on RSU vesting). Sell-side coverage skews bullish: Guggenheim initiated Buy at $12 PT, Oppenheimer Outperform at $10 PT, HC Wainwright Buy — vs. ~$1.80 spot, that’s a 5-7x implied upside if you believe the consensus, which I tend not to.

The Science: Mechanism & Chemistry

All three assets are biologics — recombinant monoclonal antibodies.

• LOQTORZI (toripalimab) is a humanized IgG4κ anti-PD-1 antibody differentiated by its binding to the FG loop of PD-1, distinct from the C’D loop binding of Keytruda and the N-terminal loop of Opdivo.

• Casdozokitug (CHS-388/SRF388) is a recombinant human IgG1 monoclonal antibody — the only clinical-stage IL-27 antagonist known.

• Tagmokitug (CHS-114/SRF114) is an afucosylated human IgG1 anti-CCR8 antibody designed for enhanced ADCC against CCR8-expressing tumor-infiltrating Tregs.

None of these are first-in-class targets — PD-1 is a 7-drug shark tank — but tagmokitug and casdozokitug have first-in-class mechanism positioning within their narrower target classes.

Mechanism Validation. PD-1 is arguably the most validated immuno-oncology target in history; toripalimab is approval #8 in the class. CCR8 is mechanistically de-risked through extensive Treg biology but no anti-CCR8 antibody has been approved — BMS-986340 (Bristol-Myers), GS-1811 (Gilead/Jounce), AbbVie ABBV-514, Shionogi S-531011, Bayer BAY 3375968, Roche RG6292, Amgen, Sino Biopharm, and Nanjing Immunophage are all in clinical development. IL-27 is the speculative one — casdozokitug is the only IL-27 antagonist in the clinic, and the target has no clinical validation outside of casdozo’s own data. That’s a double-edged sword: pristine optionality if it works, but no fallback if the mechanism doesn’t translate.

Manufacturing / CMC Risks. A real one. Coherus is “executing a coordinated initiative to onshore the biomanufacturing process for LOQTORZI, casdozokitug and tagmokitug to the United States”. Translation: Junshi makes toripalimab in China, and BIOSECURE Act exposure is a live issue. The Company acknowledges this is a multi-year transition with FDA approval pending for the US drug substance manufacturer, which was achieved for LOQTORZI in Q1 2026 (allowing them to begin capitalizing US-made inventory — note the $9.7M non-current inventory build in Q1). They’ve explicitly flagged risk that “we may not be able to achieve regulatory approval for the initiative or the full strategic benefits expected to result from the onshoring initiative”. Single-source CMOs across all three assets is a chronic chokepoint.

Biochemical Deep Dive

The Target

Three different bets across the immune escape continuum.

• PD-1 is the canonical T-cell checkpoint — binding to PD-L1/L2 on tumors and APCs delivers an inhibitory signal that exhausts effector T cells; blockade reactivates them.

• IL-27 is a heterodimeric IL-12 family cytokine (p28/EBI3 subunits) secreted primarily by tumor-associated macrophages and antigen-presenting cells, and it’s wildly overexpressed in HCC, RCC, and NSCLC. It signals through IL27Rα/gp130 on multiple immune cells (T, NK, macrophages) and exerts net immunosuppression — it’s elevated during pregnancy to maintain maternal-fetal tolerance, which is essentially a feature analog of tumor immune escape.

• CCR8 is a G-protein-coupled chemokine receptor selectively expressed on the most suppressive subset of tumor-infiltrating Tregs (CD4+ FOXP3hi). It’s almost absent on peripheral Tregs or other T cell populations, making it one of the cleanest known marker for tumor-resident Tregs. The biological rationale for selective tumor-Treg depletion has been built over a decade of preclinical work in colorectal, breast, and lung cancer models.

The Chemistry

• LOQTORZI’s receipts-level differentiator is FG loop binding — Junshi engineered toripalimab to bind a different epitope than Keytruda or Opdivo, with claimed differences in glycosylation site dependence and Fc-effector profile. In practice, no head-to-head data exists, and clinical efficacy in NPC looks comparable to other PD-1s in the same setting.

• Casdozokitug is a more interesting molecule. It’s a fully human IgG1 designed to bind IL-27 itself (the heterodimer) and block its interaction with IL27Rα, sparing other IL-12 family cytokines (IL-12, IL-23, IL-35). The IgG1 backbone with retained Fc effector function is a deliberate design choice — preclinical data suggest casdozo can deplete some IL-27-producing macrophages via ADCP, in addition to neutralizing soluble cytokine.

• Tagmokitug is the cleanest piece of molecular engineering of the three. It’s an afucosylated IgG1 — meaning fucose has been removed from the N-glycan on the Fc region, which increases binding to FcγRIIIa on NK cells by ~50-fold and amplifies ADCC potency. This is the same engineering approach used in mogamulizumab (Poteligeo), trastuzumab biosimilar Margenza, and obinutuzumab. The afucosylation is not novel chemistry, but applying it to CCR8 Treg depletion is a coherent design choice — you want cytolytic depletion, not antagonism, since Tregs are abundant and replenishable. The competing CCR8 antibodies (BMS-986340, GS-1811, etc.) are also Fc-enhanced for ADCC, so this isn’t a clear best-in-class differentiator on its own.

The Mechanism

Casdozokitug’s mechanism in HCC, per Coherus, is to relieve IL-27-mediated suppression of effector T cells and NK cells while preserving Th1 polarization and IFN-γ production. The combination rationale with PD-1 blockade (toripalimab) plus VEGF blockade (bevacizumab) attacks three layers: vascular normalization (bev), checkpoint inhibition (tori), and cytokine-mediated suppression release (casdozo). The triplet hypothesis is that you depthen and durabilize the doublet response.

Tagmokitug’s mechanism is more direct: ADCC-mediated depletion of CCR8+ intratumoral Tregs, releasing CD8+ T cells from suppression. Combination with PD-1 blockade is required because depleting Tregs alone gives the unleashed T cells fuel but not direction — anti-PD-1 reactivates them while tagmokitug removes the brake on the brake.

The Biomarker Receipts

This is where Coherus actually has the cleanest story. The SITC 2025 multiomic biomarker package for tagmokitug — across n=13 HNSCC patients with paired tumor biopsies — showed: 74% reduction in tumor CCR8+ Treg density, 43% reduction in total FOXP3+ Treg density, 73% increase in CD8+ T cell density, and a 12-fold improvement in the CD8+/CCR8+ Treg ratio. Non-CCR8 Tregs were spared, confirming selectivity. CD4+ and CD8+ T cells in peripheral blood were preserved.

That’s a real proof-of-mechanism package — turning “cold” tumors “hot” with measurable, statistically significant biomarker changes is harder to fake than ORR in small studies. It establishes that the molecule is doing what it was designed to do; it doesn’t establish that this will translate to durable clinical benefit, which is the gap that has killed many mechanism-validated I/O programs (e.g., LAG-3, TIGIT). For casdozokitug, the biomarker work is less robust — Coherus has shown IL-27 pathway suppression and downstream IFN-γ activation signatures from the Phase 1 work but the Phase 2 HCC data is primarily efficacy-driven, not biomarker-driven.

Bottom Line for Biology. Tagmokitug has the cleanest mechanism-of-action receipts of any non-BMS CCR8 program, which makes it strategically attractive as either a standalone Phase 2 asset or M&A bait. Casdozokitug’s biology is more speculative because the IL-27 target has no clinical precedent — the mid-2026 randomized Phase 2 readout is the first time anyone will know if IL-27 antagonism is real medicine.

Clinical Data

Efficacy

LOQTORZI in NPC (JUPITER-02): At 6-year follow-up presented at ESMO Asia 2025, median OS for the toripalimab + GP arm was 64.8 months vs. 33.7 months for placebo + GP (HR 0.62, 95% CI 0.45–0.85). Median PFS at the final analysis was 17.3 vs. 8.1 months (HR 0.43); ORR was 78.8% vs. 67.1% (p=0.022); median DOR was 18.0 vs. 6.0 months. These are franchise-defining survival numbers for r/m NPC, but the addressable patient pool is tiny — GLOBOCAN reports ~2,000 new US cases/year, of which only a fraction become r/m and eligible.

Casdozokitug in 1L HCC (Phase 2, single-arm with atezolizumab + bevacizumab): At ASCO-GI 2025 final readout in 29 patients, the combo delivered ORR 38%, CR rate 17.2%, mPFS 8.1 months. Cross-trial caveat (and it’s a big one): the IMbrave150 atezo + bev benchmark in roughly comparable 1L unresectable HCC populations was ORR ~30%, CR rate ~7–8%, mPFS 6.9 months, mOS 19.2 months. If casdozo’s 17.2% CR and 8.1-month PFS hold up under randomized comparison, that’s a meaningful incremental benefit — but n=29, single-arm, single-cohort can show artifactual response inflation through patient selection bias, more aggressive imaging cadence, or central read differences. The 17% CR figure is roughly 2.5x the IMbrave150 historical control rate, which is suspicious-large for an add-on mechanism without a stronger preclinical rationale. The competing 1L HCC space includes STRIDE (durvalumab + tremelimumab), with ORR 20.1%, mOS 16.4 months, 48-month OS rate 25.2% and the newly emerging anlotinib + penpulimab (APOLLO trial 2025).

Tagmokitug in 2L+ HNSCC (Phase 1 expansion): One confirmed partial response in a heavily pretreated PD-1 refractory patient at AACR 2025. In context: the standard-of-care benchmark for 2L PD-1-refractory HNSCC is roughly ORR 13–14% with cetuximab or single-agent chemotherapy, with mPFS ~2–3.8 months. Nab-paclitaxel + nivolumab in PD-1 refractory R/M HNSCC has shown ORR ~46% in small studies. A single PR in n≈21 is signal-detection level, not signal — it’s the data-floor before the Phase 1b dose optimization mid-2026 readout.

The P-Hacking Check

The casdozo Phase 2 ORR climbed during follow-up: initial readout was 27% ORR with 10.3% CR, then “increased to 38% ORR and 17.2% CR” at the final readout. ORR maturation over time is biologically real (responses deepen with continued treatment), but the magnitude of the swing — 11 percentage points on ORR, 7 percentage points on CR — in a 29-patient trial is data the bears will hammer on. The denominator stayed at 29; this isn’t denominator games. It is, however, almost entirely dependent on assessments in a tiny trial, where 2-3 additional CRs move the headline materially. The randomized Phase 2 readout will use blinded independent central review on 72 patients — that’s where p-hacking artifacts wash out.

Safety/Tolerability — The Quiet Killers

• LOQTORZI’s class label includes immune-mediated reactions: pneumonitis, hepatitis, colitis, endocrinopathies, nephritis, dermatologic toxicities, plus rare cases of myocarditis and Guillain-Barré syndrome. No drug-specific Black Box; standard PD-1 monitoring requirements apply. JUPITER-02 reported “no new safety signals” through 6 years of follow-up, which is the longest mature PD-1 safety database in NPC.

• Casdozokitug + atezo + bev safety profile per ASCO-GI 2025: “consistent with the known safety profiles of atezolizumab and bevacizumab, with no new safety signals identified”. This is the company’s framing; specific Grade ≥3 AE rates and discontinuation data were not in the headline materials I could verify. A quiet killer that often emerges in HCC immuno-combo trials is liver function deterioration (Child-Pugh class drift) and bevacizumab-related bleeding (variceal hemorrhage in cirrhotic patients with portal hypertension). Watch for these in the randomized Phase 2 readout.

• Tagmokitug: “manageable safety profile” with “no dose-limiting toxicities” reported across the 5–1200mg dose range. Treatment-related AEs in Phase 1 monotherapy were 33% all Grade 1-2, with pyrexia 13% most common — these are exactly the AE patterns you’d expect from a cytolytic IgG1 antibody and not concerning. The longer follow-up question is whether sustained Treg depletion creates autoimmunity downstream — Tregs are critical regulators of systemic immune homeostasis, and we don’t yet know how durable depletion behaves over 12+ months of dosing.

Data Integrity

JUPITER-02 was a 289-patient randomized, double-blind, placebo-controlled, multiregional Phase 3 with independent central review of PFS as primary endpoint. Pristine design.

The casdozokitug HCC Phase 2 (atezo + bev combo) was 29 patients, single-arm, open-label, single-cohort — academic-quality, not registration-quality. The randomized Phase 2 (CHS-388-202, n=72, 1:1:1) reading out mid-2026 fixes the design issue, but the absolute n is still small. If casdozo extends ORR/PFS over the tori/bev internal control, that’s a real but not registration-sufficient signal — a Phase 3 will be required.

The tagmokitug Phase 1 has been heavily biomarker-focused with required paired tumor biopsies — high-quality translational data, low-quality efficacy data (small numbers, heavily pretreated, no contemporaneous control).

Pipeline

LOQTORZI (toripalimab) — Approved.

• r/m NPC 1L combo and 2L+ monotherapy in US and Canada.

• Value driver, but capped by indication size. NCCN category 1 preferred for both 1L combo and 2L+ monotherapy.

• Q1 2026 net revenue $11.8M, +61% YoY; full-year 2025 was $40.8M.

• Expansion potential exists into LS-SCLC (Junshi-sponsored Phase 3 with anti-BTLA combo, NCT06095583), HPV+ oropharyngeal SCC (INOVIO Phase 3 with INO-3112), ovarian (IPROC Phase 2), and HNSCC/NSCLC (STORM Therapeutics METTL3 combo).

• Royalty drag (low 20%+ to Junshi) means even at $200M+ peak revenue, this is a low-margin asset.

Casdozokitug (CHS-388, anti-IL-27) — Phase 2 randomized.

• Lead indication 1L HCC (NCT06679985); patient accrual complete; readout mid-2026.

• Orphan Drug Designation for HCC granted October 2020 — that’s 7-year US market exclusivity if approved.

• Adimab royalty obligations are low-to-mid single-digit; Adimab also entitled to up to $10.5M per licensed product in remaining milestones.

• Value driver: high, if readout works; zero, if it doesn’t. Pipeline-defining binary asset.

Tagmokitug (CHS-114, anti-CCR8) — Phase 1b/2a.

• Multiple cohorts: 2L HNSCC and 2L upper GI dose optimization (mid-2026 readout); 1L/2L ESCC (2H 2026); 4L+ MSS CRC (2H 2026); mCRPC + pasritamig (Janssen-supplied; initiates fall 2026).

• Vaccinex royalty is low single-digit on global net sales of approved products. Vaccinex is eligible for $2M in remaining clinical milestones and $11.5M in regulatory milestones per Vaccinex Licensed Product.

• Value driver: medium-to-high optionality. Real M&A appeal here if data extends — CCR8 is in active development across most major immuno-oncology shops.

Pipeline Verdict. Casdozo carries 70% of the equity value; tagmokitug carries 25%; LOQTORZI provides the cash flow bridge but isn’t a multibagger asset given the royalty structure. The pasritamig combo is a $0 NPV optionality bet that costs nothing to run. The Junshi-licensed CHS-006 (anti-TIGIT) and the IL-2 cytokine option are both inactive — Coherus terminated the TIGIT program in January 2024.

Intellectual Property & The Moat

The summary provided below is based on the 10-K filed by the Company in March 2026 and the 10-Q filed in May 2026.

Coherus states only that “our patent portfolio includes issued or pending claims directed to formulations, methods of manufacturing biological proteins, and drug products and devices, including their methods of use and methods of manufacture.” For a company that touts pristine scientific differentiation, that’s the equivalent of “trust me, bro.”

Asset-Specific Patent Runways

LOQTORZI (toripalimab): Coherus does NOT own the composition-of-matter IP — Junshi Biosciences does. Coherus’s rights flow from the February 2021 Collaboration Agreement, which is “exclusive rights to LOQTORZI in the United States and Canada” plus options on Junshi’s anti-TIGIT (terminated) and IL-2 cytokine pipeline. The Canada license to Apotex runs until 10 years after first Canada commercial sale (January 2026), so through January 2036, with a 10-year extension option. Regulatory exclusivity backstop: LOQTORZI is a biologic and is entitled to 12-year BPCIA exclusivity from FDA approval (October 27, 2023), running through October 2035 before biosimilar applications can become effective. The composition patents — owned by Junshi — were filed circa 2014–2015 by Junshi (toripalimab was first disclosed in Junshi’s 2018 IPO documentation), so US composition-of-matter is likely good through ~2034–2035 absent patent term extension. The structural weakness here is contractual: Coherus pays Junshi a low-20%-range royalty plus up to $355M remaining in milestones — Coherus is effectively a US/Canada commercial agent, not the IP owner.

Casdozokitug (CHS-388, formerly SRF388): In-licensed via the September 2023 Surface Oncology acquisition. The original IP was generated under the Adimab platform license — the Adimab agreement provides Coherus a worldwide, royalty-bearing, sublicensable license, with Adimab assigning patents covering the selected antibodies upon exercise of Commercialization Options. Royalties to Adimab are tiered low-to-mid single-digit on worldwide net sales. Regulatory exclusivity: HCC Orphan Drug Designation granted October 2020 provides 7-year US orphan exclusivity post-approval. Standard 12-year BPCIA applies for any biologic indication.

Tagmokitug (CHS-114, formerly SRF114): In-licensed via Surface Oncology’s March 2021 license from Vaccinex covering certain antibodies, including the anti-CCR8 antibody. Worldwide, exclusive, sublicensable. Coherus pays Vaccinex up to $11.5M in regulatory milestones per Vaccinex Licensed Product plus low single-digit royalties on global net sales. The composition patents originated in Vaccinex’s drug-discovery platform; specific expirations are not disclosed.

Ownership & Licensing Structure

All three lead assets are in-licensed, not wholly owned. The royalty stack is heavy: low 20%+ to Junshi on LOQTORZI, low-to-mid single-digit to Adimab on casdozo, low single-digit to Vaccinex on tagmokitug. Layer the $355M in remaining Junshi milestones, $10.5M in Adimab regulatory milestones per product, and $13.5M aggregate remaining Vaccinex milestones plus $7.2M to Memorial Sloan Kettering for tagmokitug, and Coherus has substantial cumulative milestone exposure relative to its $250M market cap.

Competitive Landscape

LOQTORZI faces 8 approved anti-PD-1/PD-L1 antibodies in the US (Keytruda, Opdivo, Tecentriq, Imfinzi, Bavencio, Libtayo, Jemperli, TEVIMBRA) plus Akeso’s penpulimab approved April 2025 in NPC — directly competitive. The $355M Junshi milestone stack means even ramping LOQTORZI revenue triggers payment obligations that crimp operating leverage.

Casdozokitug as the only clinical-stage IL-27 antagonist has an open field, but the 1L HCC SOC bar is high: atezolizumab + bevacizumab (Roche, ~$1.5B+ HCC franchise globally) and STRIDE durvalumab + tremelimumab (AstraZeneca, ~$500M+ HCC franchise) are entrenched.

Tagmokitug enters a CCR8 shark tank: BMS-986340 (NCT05537740), Gilead’s GS-1811 (NCT05007782), and at least 5 other Phase 1/2 programs from AbbVie, Bayer, Roche, Amgen, Shionogi, and Sino Biopharm.

The Verdict

Scientific Conviction: Medium. Tagmokitug’s mechanism-of-action receipts are pristine so far — selective intratumoral Treg depletion with a >50% CD8+ T cell uptick is a clean biology story. Casdozokitug’s 17.2% CR rate is impressive in absolute terms but requires randomized confirmation given the IL-27 target’s lack of clinical precedent. LOQTORZI is approved, validated, and unremarkable right now.

Commercial Viability: Low-to-Medium. NPC is small, Penpulimab is approved competition, and the Junshi royalty stack caps margins. HCC and HNSCC offer real franchise potential if casdozo/tagmo work — $3B+ TAMs each — but Coherus’s 13–18% operating margin ceiling post-royalties (assuming low-20% royalty + low single-digit on novel assets + the 5% revenue interest) makes standalone profitability dependent on substantial blockbuster scale.

M&A Appeal: Medium-to-High. Casdozo + tagmo is a credible portfolio for AstraZeneca (HCC franchise expansion, behind STRIDE), Roche (HCC + CCR8 platform fit), BMS (CCR8 consolidation play), Merck (Keytruda combo expansion), or Gilead (CCR8/I-O bolt-on). A bidder would likely run the deal post-mid-2026 readout to de-risk the binary. A pre-readout takeout at premium is possible but not the base case.

Trader Profile: Binary event gamblers. Mid-2026 casdozo readout dominates the equity beta. Not a long-term compounder; not a momentum trade; not for risk-averse capital. Spec-vestors looking for asymmetric pre-readout setups, options traders selling premium into the binary, and M&A spec hands who would view positive readout as activating the strategic clock.

BUY Thesis

Target Audience: Specialty biotech investors with high binary-risk tolerance, options traders comfortable with vega/theta capture, and event-driven hedge funds positioning for the mid-2026 casdozo readout.

Rationale: At ~$250M market cap with $167M cash, the enterprise value is ~$80M for an asset portfolio that includes a $40M+ revenue commercial product, a Phase 2 randomized asset (casdozo) reading out in 6–8 weeks with credible single-arm precedent (17.2% CR vs. ~7–8% historical), a mechanism-validated CCR8 program (tagmokitug) with multiple cohorts reading out 2H 2026, free optionality on the Janssen pasritamig combo in mCRPC, and up to $75M in UDENYCA earnouts. The sell-side target average is $9–12 (5-7x upside). Asymmetric.

Execution/Strategy: Consider building position in tranches ahead of the mid-2026 readout — consider initial position 50% of intended size now, layering 25% over the next 4 weeks, holding final 25% as dry powder for either a pre-readout pullback or a post-readout follow-on offering. Options framing: with IV typically elevated 30–40 days ahead of a biotech binary, consider selling cash-secured OTM puts at the $1.50 strike (~25–30% delta) to capture theta and vega decay while improving cost basis. If you want defined-risk binary exposure, the $2 / $3 call vertical for the July 2026 expiry will price in the readout. Avoid naked long calls into a 50/50 binary — IV crush usually wins.

HOLD / WATCH LIST Thesis

Target Audience: Generalist healthcare PMs, drug-development specialists who want to underwrite the data before committing capital, and anyone who doesn’t have an edge on what 72 randomized patients will produce.

Rationale: The mid-2026 readout is unknowable with confidence. The casdozo Phase 2 single-arm data is compelling but small. The Junshi royalty structure caps the bull case. Insider activity is muted, and the absence of specialist biotech anchor funds in the 13G/13F filings is a yellow flag — smart money usually shows up before high-conviction binaries. Consider waiting for the data; if positive, you could pay 50–100% more, but you’ll have certainty on the central scientific question and a clearer path to franchise economics. If negative, you could save 50–80% of your capital.

Execution/Strategy: Consider setting price alerts at $1.20 (re-entry below the secondary offering at $1.75 minus a discount) and $2.50 (post-readout pop confirmation level). Track casdozo updates for any pre-readout signal. If management announces a raise via the ATM or a marketed deal at $1.50–$1.70, consider taking that as evidence of management’s own confidence in pre-readout valuation.

SELL / AVOID Thesis

Target Audience: Risk-averse capital, ESG funds skittish about binary biotech, and traders looking for trends rather than events.

Rationale: Pre-revenue at scale, structurally royalty-taxed on the lead commercial asset, dependent on a 72-patient Phase 2 binary reading out in 6–8 weeks, run by a CEO (Dennis Lanfear) who has demonstrated willingness to dilute below market price (the Feb 2026 offering at $1.75 came at a significant discount to the prior ~$2.20 trading range). Cross-trial comparison artifacts in HCC are common — LEAP-002 and COSMIC-312 both failed against atezo + bev. Tagmokitug’s “promising” data is one PR in 21 patients. The CCR8 shark tank means even positive Phase 2 data may not differentiate strongly enough against BMS-986340, GS-1811, or AbbVie’s program. Sell-the-news risk is real even on positive readout if the magnitude doesn’t blow out expectations.

Execution/Strategy: If short: consider covering any short position ahead of binary — gamma risk on positive readout is unbounded. If long and sitting on a profit from the recent rally (stock is up ~140% from June 2025 lows), consider trimming 30–50% pre-readout to lock gains, letting the rest ride for binary outcome. If neutral: consider skipping the name; there are better risk/reward setups in the small-cap biotech universe where catalyst timing aligns with cash runway.

Final Verdict

WATCH LIST. Mid-2026 readout is the entire thesis. Consider scaling in ahead of binary only with explicit binary-event risk tolerance and position sizing that survives a 60% drawdown. However, consider waiting for the data, accepting the 50–100% upside cost of certainty.

This report is strictly for informational and educational purposes only. It does not constitute financial advice, investment advice, or a recommendation to buy or sell any securities mentioned.

The scientific and clinical analyses herein should not be interpreted as medical guidance, diagnostic information, or treatment recommendations.

At the time of writing, the author does not hold a position in Coherus Oncology (CHRS).

Biotech investing is inherently volatile. Past scientific validation does not guarantee future clinical or regulatory success. Treat all clinical-stage biopharma allocations accordingly.