Command Palette
Search for a command to run...
Tolinapant in relapsed/refractory PTCL: an IAP antagonist combination with no tolerable dose
Taiho terminated its Phase 1-2 study of the IAP antagonist tolinapant plus decitabine and cedazuridine in relapsed or refractory peripheral T-cell lymphoma after no safe and tolerable combination dose could be identified. The stop was an on-mechanism tolerability ceiling, not a target-validity failure.
Mechanism Risk Score
| Component | Points |
|---|---|
| Phase-weighted failure burden | 4.7 / 40 |
| Archetype severity | 8.5 / 25 |
| Temporal recency | 4.3 / 15 |
| Genetic evidence deficit | 7.5 / 15 |
| Programmatic saturation | 2.5 / 5 |
For BIRC2 in Peripheral T-cell lymphoma, the Mechanism Risk Score is 27/100 (yellow band). The score is a failure-burden index derived from Claidex post-mortems on this target–disease pair, not a probability of approval.
MRS 27/100 (YELLOW). 1 programs across BIRC2 have been documented for BIRC2 in Peripheral T-cell lymphoma: 0 Phase 3, 0 Phase 2, 1 Phase 1 — of which 0 were efficacy failures, 1 safety, 0 biomarker, and 0 operational (enrollment, sponsor, or funding). The most informative failure on file is Tolinapant in relapsed/refractory PTCL: an IAP antagonist combination with no tolerable dose. This score quantifies the documented failure burden; the Open Targets association score of 0.50 reflects moderate genetic support, neither rescuing nor compounding the failure record. The MRS is not a prediction of future trial outcomes — it is a structured summary of the empirical record, recomputed live from the Claidex claims table, and intended to flag mechanisms where any new program must explicitly resolve each prior failure mode before pursuit is justified.
This score does not predict whether the next trial will succeed. It flags how heavy the documented mechanistic failure record is before a new program is justified.
What was tried
Tolinapant (ASTX660, ChEMBL4173974) is an oral, non-peptidomimetic small-molecule antagonist of the cellular inhibitor of apoptosis proteins cIAP1 and cIAP2 (genes BIRC2 and BIRC3) and X-linked IAP, with a ChEMBL maximum phase of 1 (ChEMBL, 2025). NCT05403450, run as ASTX660-03 by Taiho Oncology, was a Phase 1-2 open-label study of tolinapant combined with oral decitabine and cedazuridine, and of decitabine and cedazuridine alone, in subjects with relapsed or refractory peripheral T-cell lymphoma (ClinicalTrials.gov NCT05403450). The Phase 1 stage characterized safety, dose-limiting toxicities, and pharmacokinetics, and the Phase 2 stage was to read overall response rate by the 2014 Lugano classification. Actual enrollment reached 33 participants. Taiho terminated the study with the posted reason that no safe and tolerable dosing for the combination of tolinapant and decitabine and cedazuridine was identified based on protocol-defined criteria (ClinicalTrials.gov NCT05403450). The primary completion date was December 16, 2024.
The biological hypothesis
IAP proteins block apoptosis by restraining caspase activity and by shaping tumor necrosis factor receptor signaling. Antagonizing cIAP1/2 and XIAP shifts TNF-alpha signaling toward caspase-8 activation and cell death, and in T-cell lymphoma models tolinapant also drives necroptosis and markers of immunogenic cell death, giving the class both a direct cytotoxic and an immune-priming rationale (Blood Advances, 2021). A separate report showed that ASTX660 and TNF-alpha synergistically induce necroptosis, reinforcing the TNF-dependent mechanism (Biochemical and Biophysical Research Communications, 2022). Preclinical work then showed that pairing tolinapant with a DNA hypomethylating agent enhanced immunogenic potential, which was the direct rationale for combining it with decitabine and cedazuridine (Cancer Research Communications, 2024). BIRC2 falls in a small-molecule tractable class, with an advanced clinical bucket and high-quality ligand and pocket annotations in Open Targets (Open Targets Platform, 2025). Relapsed or refractory PTCL carries a high unmet need and short survival after standard salvage, so a combination that couples apoptosis restoration to epigenetic immune priming was a defensible bet.
What actually happened
The combination did not clear the tolerability bar needed to define a recommended dose. Across 33 enrolled participants the sponsor closed the study per protocol because no safe and tolerable dose of tolinapant plus decitabine and cedazuridine met protocol-defined criteria, and no efficacy stage was reached under the combination (ClinicalTrials.gov NCT05403450). IAP antagonism carries an on-mechanism liability. Because the class works through TNF-alpha-dependent cell death and cytokine release, systemic inflammatory and hepatic toxicities scale with exposure, and stacking a hypomethylating agent adds overlapping myelosuppression and further immune activation. The pharmacovigilance footprint is minimal, since tolinapant remains investigational, with a single report for TOLINAPANT in OpenFDA FAERS at retrieval, which is far too sparse to support any disproportionality analysis (OpenFDA FAERS, retrieved 2026-07-11). Tolinapant monotherapy had shown preliminary single-agent activity in relapsed or refractory PTCL and cutaneous T-cell lymphoma in the separate ASTX660-01 study, which sharpened the contrast. The molecule was active alone, but the combination window was too narrow to dose.
Failure mechanism, best guess
The proximate failure archetype was a safety signal, specifically a combination tolerability ceiling rather than a target-validity failure. The mechanistic tension was additive on-mechanism toxicity. Tolinapant releases TNF-alpha-driven death signaling and inflammatory cytokines, and decitabine and cedazuridine contribute myelosuppression plus their own immunostimulatory effects, so the two agents compress the therapeutic window from both the hematologic and the inflammatory side. Open Targets records no direct BIRC2 to PTCL association, so under the scoring convention an unavailable association defaults to 0.5, which enters the genetic-deficit term (Open Targets Platform, 2025). That absence is itself informative. The case for IAP antagonism in PTCL rested on preclinical apoptosis and immune biology, not on human genetic support for the target in this disease. When a program leans on mechanism rather than genetics, the dose-finding step carries more of the risk, and here the combination could not reach a tolerated, active exposure.
How to prevent this next time
Two quantitative tools would have tightened the read before committing a combination cohort. First, a Bayesian toxicity-driven dose-finding design with an explicit target toxicity probability, such as a BOIN or continual reassessment scheme, would have modeled the combination window directly rather than escalating on fixed rules, evaluating the predictive probability of identifying a tolerated active dose at each cohort:
With a prior informed by single-agent tolinapant exposures and the known myelosuppression of decitabine and cedazuridine, the predictive probability of a tolerated combination dose would have stayed low until early cohorts showed non-overlapping toxicity, which argues for a staggered lead-in rather than simultaneous full-dose escalation. Second, a historical base-rate adjustment should have anchored expectations, since oncology carries a Phase 1 likelihood of approval near 3.4 percent against an all-indication average near 13.8 percent (Wong, Siah, and Lo, 2019), and combination tolerability failures are a common early exit for immune-priming agents. A biomarker-guided expansion, enriching for tumors with intact TNF and caspase-8 signaling where the combination should add most, would have raised the chance that any tolerated dose also produced responses. The single highest leverage change would have been ...
What this means for similar programs
IAP antagonism remains an active mechanism, with xevinapant, birinapant, LCL161, and APG-1387 among the class, and the lesson is that on-mechanism cytokine and apoptosis toxicity travels into every combination. Programs pairing an IAP antagonist with a hypomethylating agent or another myelosuppressive backbone should treat the tolerated combination dose as the primary uncertainty, not the single-agent activity, and should design lead-in and staggering to protect the window. Single-agent signals do not guarantee a combinable molecule, and the decision to close on tolerability rather than push an unsafe dose was consistent with that reality.
Open questions
What exposures were reached before the combination was deemed intolerable, and were the dose-limiting events hepatic, inflammatory, or hematologic. Would a staggered schedule or a lower fixed decitabine and cedazuridine dose have opened a tolerated window while preserving immunogenic priming. Does tolinapant monotherapy, which showed single-agent activity, offer a better development path in PTCL than combinations. Would a different epigenetic partner with less overlapping myelosuppression combine more safely with IAP antagonism.
Sources
- ClinicalTrials.gov, NCT05403450 (ASTX660-03), terminated status, whyStopped text, enrollment 33, Phase 1-2 design and endpoints, primary completion 2024-12-16: https://clinicaltrials.gov/study/NCT05403450 - ChEMBL, tolinapant CHEMBL4173974, small molecule, maximum phase 1: https://www.ebi.ac.uk/chembl/g/#search_results/all/query=tolinapant - Open Targets Platform, BIRC2 (ENSG00000110330) small-molecule tractability and BIRC2 to mature T-cell and NK-cell non-Hodgkin lymphoma, no direct association recorded: https://platform.opentargets.org/target/ENSG00000110330 - OpenFDA FAERS, TOLINAPANT, 1 total report, retrieved 2026-07-11: https://api.fda.gov/drug/event.json?search=patient.drug.medicinalproduct:%22TOLINAPANT%22 - Antagonism of inhibitors of apoptosis proteins reveals a novel, immune response-based therapeutic approach for T-cell lymphoma. Blood Advances.