Does tolinapant plus pembrolizumab sensitize advanced solid tumors to immunotherapy?

A phase I test of dual IAP and PD-1 blockade

The trial asked whether blocking inhibitor of apoptosis (IAP) proteins can make tumors more visible to the immune system and thereby improve the effect of pembrolizumab, a PD-1 checkpoint inhibitor. IAP proteins cIAP1, cIAP2 and XIAP suppress RIPK1-driven immunogenic cell death, a form of cell death that releases inflammatory signals and can help T cells recognize tumor cells. Tolinapant is a non-peptidomimetic antagonist designed to inhibit all clinically relevant IAPs and lower the TNF cytotoxicity threshold, so that tumor cells become more susceptible to T cell-mediated killing.

ASTEROID (NCT05082259) enrolled 23 patients with advanced cancer who had exhausted standard treatment options and had performance status 0-1. Tolinapant was given orally once daily for 7 days, followed by intravenous pembrolizumab 200 mg on day 8 of each 21-day cycle. The primary endpoints were the recommended phase 2 dose (RP2D) and safety; anti-tumor activity and pharmacokinetics were secondary endpoints. Eighteen patients completed at least two cycles and were evaluable for response by immune-modified RECIST, and paired tumor biopsies before and on treatment were available from 17 response-evaluable patients for pharmacodynamic analysis.

No dose-limiting toxicities, but more dose changes at 180 mg

No grade 4 or 5 treatment-related adverse events occurred, and no dose-limiting toxicities were identified during the 6-week observation period. The most frequent treatment-related events were grade 1-2 rash (43%), alanine aminotransferase increase (43%), lipase increase (30%), aspartate aminotransferase increase (30%) and nausea (26%). Grade 3 events included lipase elevation (13%) and single cases each of nausea, eosinophilia and lymphopenia, all described as transient.

The Bayesian TITE-CRM model set the maximum tolerated dose at 180 mg (dose level 4), with a dose-limiting toxicity rate of 0.030 (90% credible interval 0-0.222). Yet dose modifications due to treatment-related events were needed in 5 of 23 patients overall (22%) and in 3 of 7 patients at 180 mg (43%). Because biological activity, efficacy and tolerability were weighed together, the safety review committee chose 150 mg (dose level 3) as the RP2D rather than the model-defined maximum tolerated dose. This distinction matters: the highest dose that avoids protocol-defined toxicity is not automatically the dose carried forward.

A 33% response rate concentrated in ICB-resistant tumors

The overall response rate was 33%, and confirmed responses appeared across dose levels 2 to 4. Responses included three patients with breast cancer (two ER+/HER2- and one PD-L1-negative triple-negative), one cervical adenocarcinoma, one oral squamous cell carcinoma and one pancreatic adenocarcinoma. In the discussion the authors describe partial responses in six patients whose tumors are typically unresponsive to pembrolizumab alone, and note that two of ten patients with ER+ breast cancer achieved a partial response at dose level 2.

These numbers come from a small, heavily pretreated population: the median number of prior systemic treatments for metastatic disease was 5 (interquartile range 2-6), and 4 of 23 patients (17.4%) had prior checkpoint inhibitor treatment. The 33% figure is therefore a proportion of 18 response-evaluable patients in a single-arm phase I study, not a comparison against pembrolizumab alone. It is early evidence of activity, and the authors frame it as clinical proof-of-concept rather than as an established efficacy estimate.

IAP blockade was achieved but did not track with response

Pharmacodynamic sampling showed that oral tolinapant reached systemic exposure at all dose levels tested. cIAP1 depletion was confirmed in blood and in tumor cells by immunohistochemistry, with sustained suppression across the 7-days-on/14-days-off cycle; the authors attribute the persistence to ongoing cIAP1 degradation rather than prolonged drug exposure, since the mean half-life at 180 mg was 18 hours. Downstream evidence of ripoptosome assembly and caspase-8 activation was seen in tumor cells, and doses at or above 120 mg were considered biologically active on this basis.

A key negative finding limits how the drug can be monitored: neither baseline cIAP1 expression nor the extent of its depletion in peripheral blood mononuclear cells or tumor samples correlated with clinical response. Target engagement confirms that the drug hit its intended protein, but it does not identify who benefited. The authors also note that cIAP2 and XIAP engagement could not be confirmed in patient samples because the assays were not adequately validated, so the claim of balanced pan-IAP activity rests partly on prior preclinical work.

Response tracks with a TNF/IFN-enriched tumor microenvironment

Integrated translational analyses compared responders and non-responders using bulk RNA sequencing of tumor biopsies, plasma protein profiling and T cell receptor sequencing. Response was associated with a TNF/IFN-enriched inflammatory tumor microenvironment, increased tumor antigenicity and enhanced TCR clonal diversity. In other words, the tumors that shrank already carried, or developed, the inflammatory and antigen-presenting features needed for T cells to recognize them.

Resistance showed a different pattern: a CCR4/6/7/8-driven niche dominated by tumor-associated macrophages and regulatory T cells. The paper presents these as associations between molecular features and clinical outcome in a small cohort, not as proven causal mechanisms. They are candidate biomarkers for stratifying patients in future studies, and the authors explicitly describe them that way rather than as validated predictive tests.

Immune-related toxicity accompanied clinical benefit

Pneumonitis occurred in four patients treated at dose levels 3 and 4, and all four experienced clinical benefit. The authors state that the study was not powered to dissect this relationship and offer several plausible explanations: pneumonitis may reflect on-target systemic immune activation in patients mounting the strongest anti-tumor responses; patients may have had pre-existing inflammatory conditions predisposing to both response and toxicity; or enhanced non-canonical NF-kB-driven inflammatory signaling may lower the threshold for inflammatory damage in normal tissue such as lung. Three of the four cases had predisposing risk factors such as extensive lung metastases or prior chest radiotherapy, although the pneumonitis areas did not radiologically overlap with metastases or prior treatment fields.

The authors also compare toxicity across settings: a phase 2 tolinapant monotherapy trial using 180 mg on a more intensive 7-days-on/7-days-off schedule reported lower rates of rash (24%) and pneumonitis (3%), while pembrolizumab monotherapy pneumonitis rates are typically 1-5%. They conclude that the combination enhances anti-tumor activity but also increases immune-mediated toxicities, particularly rash and pneumonitis. This is a hypothesis-generating observation from a small trial, not a demonstrated link between response and toxicity.

Single-arm evidence with a small response-evaluable group

The trial has no randomized control arm, so it cannot show that the combination outperforms pembrolizumab alone; the comparison to typical checkpoint-inhibitor resistance is contextual, drawn from the tumor types enrolled. Response evaluation required at least two cycles of treatment, which excluded three patients who progressed within 6 weeks, and the response-evaluable group was 18 of 23 enrolled patients. Translational comparisons between responders and non-responders rest on unpaired groups of limited size, and the biomarker findings are associations rather than validated predictive rules.

The dose-finding also carries a practical caveat: the maximum tolerated dose of 180 mg required frequent dose modifications, which is why 150 mg became the RP2D. Any future efficacy testing will use the lower dose, so the activity observed across dose levels 2 to 4 should not be read as evidence that 180 mg is preferable.

The RP2D choice, not the response rate, is the trial's clearest decision

The most consequential result for future work is the separation between the model-defined maximum tolerated dose and the dose actually carried forward. Tolinapant at 180 mg was tolerable by protocol criteria, yet 43% of patients at that level needed dose modifications, so the safety review committee selected 150 mg with pembrolizumab for expansion. That choice, together with the observation that cIAP1 depletion depth did not predict response, means the field should not treat target engagement as a surrogate for benefit.

The biomarker split is the second priority: a TNF/IFN-enriched, antigen-rich, TCR-diverse microenvironment marked response, while a CCR4/6/7/8-driven macrophage and regulatory T cell niche marked resistance. Future stratification should test those features prospectively rather than assume that IAP blockade alone identifies likely responders. The ER+ breast cancer responses are the signal most worth following, because that subtype is where checkpoint blockade usually fails and where a sensitizing combination would matter most.

Sources used in this answer