Best of ESMO: pan-tumor highlights

The phase II ROME trial, which was presented during the ESMO presidential symposium, evaluated the efficacy and safety of molecularly guided therapy versus investigator’s choice therapy in patients with progressive advanced solid tumors. Prior to randomization, molecular profiling was performed using the FoundationOne Tissue CDx and Liquid CDx assays, and the case of each individual patient was discussed in a molecular tumor board.

Druggable pathways included ERBB2 (Erb-B2 Receptor Tyrosine Kinase 2) amplification/mutations, BRAF (B-Raf Proto-Oncogene, Serine/Threonine Kinase) mutations, ALK (Anaplastic Lymphoma Kinase) translocations, RET (Rearranged During Transfection Proto-Oncogene) alterations, CDK4/6 (Cyclin-Dependent Kinases 4 and 6), CDKN2A/B (Cyclin-Dependent Kinase Inhibitor 2A and 2B) alterations, BCR-ABL (Breakpoint Cluster Region–Abelson Murine Leukemia Viral Oncogene Homolog Fusion) translocations, SMO/PTCH1 (Smoothened, Frizzled Class Receptor/Patched 1) mutations, JAK (Janus Kinase) mutations, FGFR1/2/3 (Fibroblast Growth Factor Receptors 1, 2, and 3) alterations, PI3K (Phosphoinositide 3-Kinase), AKT (AKT Serine/Threonine Kinase), PTEN (Phosphatase and Tensin Homolog), mTOR (mechanistic target of rapamycin) alterations encompassing mutations, deletions, and amplifications, NTRK1/2/3 (Neurotrophic Receptor Tyrosine Kinase 1, 2, and 3) fusions, ROS1 (ROS Proto-Oncogene 1, Receptor Tyrosine Kinase), METex14, MET alterations including amplifications, BRCA 1/2 (Breast Cancer Gene 1 and 2) mutations, and homologous recombination deficiency (HRD), for which specific inhibitors were available. Patients with a high tumor mutational burden (TMB) > 10 or mismatch repair deficiency were eligible for immune checkpoint blockade.

Notably, patients with a high TMB benefited significantly from targeted/immunotherapy (HR: 0.15; 95% CI: 0.03–0.73; p = 0.0062). However, the median overall survival (OS) was not significantly improved (9.2 months vs. 7.6 months; HR: 0.89; 95% CI: 0.68–1.13; p = 0.2998), although it is worth noting that 52% of SoC patients crossed over to receive molecularly targeted therapy. No new safety signals were detected beyond what is already known about the individual drugs [1].

While tumor-agnostic trials such as the SHIVA, MOSCATO, or NCI match trials were conducted previously with mixed outcomes, the ROME study adds to the evidence that tumor-agnostic molecular-guided therapy is feasible and improves patient outcomes [2‐4]. The ESMO recommends the use of multigene next-generation sequencing (NGS) for patients with advanced cancer to detect tumor agnostic alterations (NTRK 1/2/3 fusions, RET fusions, BRAF V600E mutations, FGFR 1/2/3 fusions, TMB high and mismatch repair deficiency; [5]). Clinical studies such as the ROME trial support its implementation in daily clinical routine.

Protein arginine methyltransferase 5 (PRMT5) modulates several key biological processes, including cell proliferation and growth, cell migration, and DNA repair [6]. Although PRMT5 is a valuable target for cancer therapy, initial trials with first-generation PRMT5 inhibitors reported considerable toxicity [7, 8].

To overcome these challenges, a new approach was developed, leading to the creation of AMG 193. This drug belongs to a new class of methylthioadenosine (MTA)-cooperative PRMT5 inhibitors that selectively target methylthioadenosine phosphorylase (MTAP)-deficient tumors. MTAP deletion, detected in 10–15% of tumors, leads to MTA accumulation in the cell, partially inactivating PRMT5. AMG 193 binds to the MTA-bound state of PRMT5, resulting in synthetic lethality [9].

During the ESMO Presidential Symposium, results from a phase I first-in-human study of AMG 193 in MTAP-deficient tumors were presented. MTAP deficiency was determined locally by detecting MTAP or CDKN2A deletion using NGS or by a central laboratory through the absence of MTAP protein expression using immunohistochemistry.

Patients in the study had received a median of two prior lines of therapy. A total of 167 patients were enrolled in dose-escalation and dose-expansion cohorts. The most common cancer types included pancreatic adenocarcinoma (PDAC; 28.7%), non-small cell lung cancer (NSCLC; 17.4%), and biliary tract cancer (BTC; 14.4%).

The safety profile of AMG 193 was more favorable compared to first-generation PRMT5 inhibitors. The most common treatment-related grade 3/4 adverse events (AEs) were nausea (4.6%), vomiting (3.4%), and fatigue (1.1%). Nausea and vomiting were manageable with anti-emetics such as ondansetron and resolved with continued dosing within 2–4 weeks. No clinically relevant myelosuppression was detected. The overall grade 3/4 AE rate was 18.4%.

Regarding efficacy, responses were observed across multiple tumor types, including difficult-to-treat cancers such as PDAC (5/23), NSCLC (5/17), and BTC (2/19). Responses appeared durable, with a median duration of response (DOR) of 8.3 months (95% CI: 2.7–not estimable; [10]).

Early data from AMG 193 appear promising; however, several questions remain unanswered, such as identifying the optimal biomarker for epigenetic modeling and determining effective combination therapies. This is of particular clinical importance, as glioblastoma multiforme—the tumor type with the highest frequency of MTAP deletion—did not show a response in this study.

The novel antibody–drug conjugate (ADC) YL201, which targets B7-H3 and carries a potent new topoisomerase‑1 inhibitor as its payload, was tested in a phase I study presented during a proffered poster session. B7-H3 is overexpressed in tumor cells, and there is only low expression in normal cells. Functionally, B7-H3 belongs to the B7 superfamily and is an immune checkpoint molecule but is also associated with cell proliferation and migration [11].

In the aforementioned trial, 312 patients with solid tumors were included in a dose-escalation and dose-expansion cohort. All patients were previously exposed to SoC therapies and 60% had already received at least two treatment lines. The most common cancer types included small cell lung cancer (SCLC) with 79 patients, NSCLC including lymphoepithelioma-like carcinomas (LELC) with 69 patients, nasopharyngeal cancer (NPC) with 75 patients, and esophageal squamous cell carcinoma (ESCC). The ORR in the expansion cohort was 44.6% and the disease control rate was 83.7%. In particular, the ORRs for SCLC (68.1%; 95% CI: 56.0–78.6), NPC (48.6%; 95% CI: 36.4–60.8), adenocarcinoma NSCLC (29.2%; 95% CI: 12.6–51.1), LELC (60.9%; 95% CI: 38.5–80.3), and ESCC (27.8%; 95% CI: 14.2–45.2) were promising. The DOR for those tumor types ranged from 3.5 months for ESCC to 11.1 months for NPC. As for the safety profile, the most common AEs were myelotoxicity, and treatment-related grade ≥ 3 AEs occurred in 51% of the patients [12].

The novel vedotin ADC, PDL1V, targets programmed death-ligand 1 (PD-L1) and was evaluated in a phase I trial involving 94 pre-treated patients with NSCLC, head and neck squamous cell carcinoma (HNSCC), breast cancer, and esophageal cancer. The study included a dose-escalation and dose-optimization cohort, followed by a dose-expansion phase for HNSCC and NSCLC patients.

The safety profile was manageable, with a treatment-related AE rate of 78.4% in the recommended phase II dose cohort of 1.5 mg/kg. The most common treatment-related grade 3 AEs included muscular weakness (2.1%), peripheral motor neuropathy, pneumonitis, pruritus, and hyponatremia (1.1% each).

In PD-L1-positive tumors (as determined by historical local testing), the ORR ranged from 12.1% to 25%, depending on the dose level administered. Notably, NSCLC patients demonstrated an ORR of 33.3%, while HNSCC patients had an ORR of 10.5%. Pembrolizumab plus PDL1V combination cohort studies are currently enrolling patients for these tumor types [13].

There is growing excitement surrounding the next generation of ADCs. However, the potential of ADCs targeting immune checkpoints such as B7-H3 or PD-L1 remains unclear. On one hand, immune-related AEs, such as pneumonitis observed with PDL1V, raise concerns about whether these effects could be exacerbated in pembrolizumab-based combination cohorts. On the other hand, no predictive biomarkers have been identified, and in the aforementioned B7-H3 trial, the activity of YL201 varied widely across tumor types.

Dysregulation of chromatin dynamics is a crucial step in cancer progression. Aberrations in subunits of the SWItch/Sucrose Non-Fermentable (SWI/SNF) family of chromatin remodeling complexes contribute to this process. Subunits such as SMARCA4 (SWI/SNF-related, matrix-associated, actin-dependent regulator of chromatin, subfamily A, member 4) and SMARCA2 (SWI/SNF-related, matrix-associated, actin-dependent regulator of chromatin, subfamily A, member 2) are frequently mutated in human malignancies. The loss of both subunits generally has synthetic lethality, and SMARCA4-deficient tumors rely on SMARCA2 activity for survival [14].

PRT3789, a first-in-class SMARCA2 protein degrader, was evaluated in a phase I trial in patients with advanced SMARCA4-deficient tumors. A total of 65 patients were enrolled, with the most common cancer types being NSCLC (52.3%), PDAC (9.2%), and breast cancer (6.2%).

Although 50.8% of patients experienced a grade ≥ 3 AE, only 4.6% of events were deemed treatment-related. The most frequent AEs included nausea (24.6%), decreased appetite (18.5%), and fatigue (18.5%). Among 46 efficacy-evaluable patients, NSCLC and esophageal cancer patients appeared to benefit most, with tumor shrinkage observed in 29% of these patients. No tumor shrinkage was observed in other tumor types [15].

The future development program for PRT3789 is currently focused on combination therapies with docetaxel and pembrolizumab, most likely in NSCLC and esophageal cancer.

For SMARCA4-deficient malignancies, which are typically aggressive tumors, no specific anti-tumor therapy currently exists. Anecdotal clinical evidence suggests that immune checkpoint inhibitors may be effective [14]. Against this background, protein degraders such as PRT3789 could represent a significant step forward in the field. However, its single-agent activity appears to be limited, making combination approaches the most likely path for effective treatment.