ABP 234 (pembrolizumab biosimilar candidate)

ABP 234 is Amgen's biosimilar candidate to Merck's Keytruda (pembrolizumab), the anti-PD-1 antibody that generated roughly $29.5B in 2024 revenue and remains the single largest-selling drug in the world [1][2]. Amgen has two Phase 3 pharmacokinetic (PK) similarity studies running in advanced non-squamous non-small cell lung cancer, both active but no longer recruiting [3][4]. The strategic clock is Keytruda's US composition-of-matter patent, which expires in 2028; whoever files a clean BLA (Biologics License Application, the FDA submission required for approval of a biologic drug) package first gets the earliest launch window into a market where a single percent of share is worth hundreds of millions per year [5]. Samsung Bioepis is currently the visible front-runner in the pembrolizumab biosimilar race, having announced positive global Phase 3 results in June 2026 and filed its first regulatory application (in Korea) in August 2026 [12].

This is not a novel compound. ABP 234 is engineered to match the amino acid sequence, glycosylation profile, and PD-1 binding characteristics of pembrolizumab closely enough to satisfy FDA biosimilarity standards. Biosimilars do not qualify for breakthrough therapy, fast track, or orphan designations because the regulatory path is a comparative BLA under the 351(k) pathway (a shortened FDA approval route created by the Biologics Price Competition and Innovation Act for biologics that are 'highly similar' to an already-approved reference product), not a de novo efficacy demonstration. Two Phase 3 trials are active and no longer recruiting: NCT06311721 in advanced or metastatic non-squamous NSCLC (n=315, comparative efficacy and PK; primary completion estimated January 2028 per ClinicalTrials.gov) and NCT06430866, a pure PK similarity study (n=163) with AUC (area under the concentration-time curve, a standard measure of total drug exposure) over 21 days as the primary endpoint, with primary completion estimated in 2026 [3][4]. Amgen has not publicly guided a specific BLA submission date, but the trial status and Keytruda's 2028 US LOE (loss of exclusivity, when patents expire and biosimilars can launch) point to a filing target in late 2026 or 2027 for the PK package, with the larger comparative NSCLC dataset supporting subsequent submissions, and an FDA decision timed to allow launch immediately post-patent [5]. Amgen ran this same playbook with Amjevita, its adalimumab biosimilar, which received the first Humira biosimilar approval in 2016 and launched in 2023 after patent settlement [6].

PD-1 is a brake pedal on T cells. When a T cell recognizes something it should attack, PD-1 on its surface can grab a matching molecule (PD-L1) displayed by another cell, and that handshake tells the T cell to stand down. Tumors exploit this by decorating themselves with PD-L1, effectively flashing a 'friendly' badge at the immune system. Pembrolizumab is an antibody that jams the brake by physically covering PD-1 so it cannot bind PD-L1, freeing T cells to attack the tumor. UniProt describes PD-1 as an inhibitory receptor on antigen-activated T cells that suppresses T-cell activation upon ligand engagement, with Open Targets scoring non-small cell lung carcinoma at 0.627 and melanoma at 0.628 for evidence weight, both among the strongest validated indications for the target. The landmark KEYNOTE-024 trial (Reck et al., NEJM 2016) established pembrolizumab's overall-survival benefit over platinum chemotherapy in first-line PD-L1-high NSCLC and anchored the drug's blockbuster trajectory [7]. For a biosimilar, mechanism validation is not the question. Pembrolizumab has demonstrated survival benefit in NSCLC, melanoma, head and neck cancer, and more than 20 other approved indications. The scientific bar for ABP 234 is instead structural and pharmacokinetic: prove the copy behaves like the original at the molecular and exposure level, and clinical efficacy is presumed to carry over.

Both Phase 3 studies are designed for biosimilarity, not de novo efficacy. NCT06430866 (n=163) is a narrow-population PK similarity study with AUC from time 0 to 21 days after first dose as the primary endpoint, the standard exposure-matching benchmark that FDA and EMA require before granting biosimilar status; primary completion is estimated in 2026 per ClinicalTrials.gov [4]. NCT06311721 (n=315) is the larger comparative trial in advanced or metastatic non-squamous NSCLC, running ABP 234 head-to-head against sourced Keytruda with AUC21d again as the primary endpoint and secondary endpoints covering efficacy, safety, and immunogenicity; primary completion is estimated January 2028 per ClinicalTrials.gov [3]. This staggered readout profile positions Amgen for a BLA submission in 2027 (based on the PK study) with the larger comparative NSCLC dataset supporting label breadth, and a potential FDA decision in 2028 aligned to the composition-of-matter LOE. This is a conservative, textbook biosimilar package. The population choice matters: non-squamous NSCLC on pembrolizumab monotherapy or combination is one of the highest-volume Keytruda indications, so if ABP 234 wins approval here, extrapolation to other pembrolizumab indications is routine under FDA biosimilar policy. Both trials are listed as active, not recruiting, meaning enrollment is complete and Amgen is now running patients through follow-up. No comparator arm concerns; the reference product is the comparator by design.

Our model estimates a 51% chance this drug is eventually approved. It starts from the historical base rate for Phase 3 drugs in this area (about 85%), then adjusts using ten facts about the trial and sponsor. What moves the number most: it is helped by more secondary endpoints than usual; it is held back by the sponsor's thin or weak approval record, weak or limited earlier-phase results, and a randomized design. The other facts land near average for this stage, so they leave the estimate roughly where the base rate put it.

Efficacy risk is minimal by design. The real risks sit downstream. First, patent litigation: Merck holds a thicket of process, formulation, and method-of-use patents that extend well past the 2028 composition LOE, and biosimilar entrants routinely face 18-36 month delays even after approval while patent disputes settle [5]. Second, formulation switching: on September 19, 2025 the FDA approved Merck's subcutaneous formulation of pembrolizumab with berahyaluronidase alfa-pmph, marketed as Keytruda Qlex; the SC formulation carries fresh patent life and a strong physician-preference argument (administration in as little as one minute) that could shift the franchise away from IV before biosimilars launch [9]. Third, commercial dilution: at least five other pembrolizumab biosimilars are in late-stage development from Samsung Bioepis (SB27, positive global Phase 3 results announced June 2026), Celltrion (CT-P51), Bio-Thera, Alvotech, Biocon, and Formycon/Zydus, and simultaneous launches will compress pricing fast [8][12]. Fourth, payer dynamics: oncology biosimilar uptake has been slower and more fragmented than immunology, with GPO (Group Purchasing Organization) contracts, 340B (a federal drug pricing program that requires manufacturers to sell outpatient drugs at deep discounts to safety-net hospitals, complicating hospital purchasing incentives) economics, and physician inertia all creating friction. Fifth, interchangeability: Amgen has not indicated that it is seeking FDA interchangeability designation, which is the higher regulatory bar that allows a pharmacist to substitute the biosimilar for the reference product without prescriber intervention. Non-interchangeable biosimilars require the prescriber to actively write the biosimilar, which historically slows uptake versus interchangeable products. Sixth, manufacturing and immunogenicity: any deviation in glycosylation or aggregation profile that shows up as a difference in anti-drug antibody rates in the comparative trial could force additional studies. None of these are unusual for the class, but stacked together they mean 85% probability of approval does not translate to 85% probability of a franchise-scale product.

Worth watching as a competitive-landscape data point, not as a standalone catalyst. Amgen's oncology biosimilar franchise (Mvasi, Kanjinti) generated meaningful but not company-defining revenue, and ABP 234 fits the same profile: reliable execution, capped upside, strategic hedge against Keytruda's post-LOE cliff. The signal to check for is BLA acceptance, which Amgen has not yet announced publicly; that filing will trigger the BPCIA patent exchange with Merck and set the litigation clock. Amgen's most recent FY2025 10-K (filed February 2026) and Q2 2026 10-Q (filed August 2026) are the right places to look for updated pipeline commentary and any patent-dance disclosures [10][11]. Samsung Bioepis appears to be the current front-runner in the class: its SB27 program reported positive global Phase 3 results in June 2026 and filed its first regulatory dossier (Korea) in August 2026, with Celltrion's CT-P51 close behind and Formycon/Zydus signaling ambition to file the first US BLA [12]. The more interesting question for BioCosm readers is not whether ABP 234 gets approved but how many pembrolizumab biosimilars reach the US market within 12 months of the 2028 LOE. If the answer is five or more launching simultaneously, price erosion will look like the Humira post-LOE experience (60-80% discount within 24 months), and no single biosimilar sponsor captures durable share. Check back after the first US biosimilar BLA acceptance is announced by any sponsor, which will set the pace for the entire field.