We are pleased to present highlights of the latest advances in immunotherapy emerging from the 2026 ASCO Annual Meeting, May 31, 2026.
(Thanks to Chinmay Jani, MD and Terri Holzen, PhD for compiling these scientific highlights)
Final safety, efficacy and biomarker results from the first-in-human study of oncolytic virus RP2 alone and in combination with nivolumab for advanced solid tumors
2504. RP2 oncolytic immunotherapy alone and in combination with nivolumab (nivo) in patients with advanced solid tumors: Final safety, efficacy, and biomarker results from the phase 1 first-in-human (FIH) study
Joseph J. Sacco (The Clatterbridge Cancer Centre and University of Liverpool, Liverpool, United Kingdom) presented the final results of a phase 1 first-in-human study of RP2 in patients with advanced solid tumors. RP2 is genetically modified HSV-1 that encodes a fusogenic glycoprotein (GALV-GP-R-), human GM-CSF, and a human anti-CTLA-4 antibody. The fusogenic glycoprotein increases immunogenic cell death by cell-to-cell fusion, and localized expression of the anti-CTLA-4 antibody promotes anti-tumor immunity without systemic immune-related toxicities. Key patient eligibility criteria for this trial included advanced or metastatic non-neurological solid tumor that had progressed on or could not tolerate standard therapy and had at least one measurable and injectable tumor >= 1 cm. Patients in the RP2 monotherapy cohort (n=25) received up to 8 doses of RP2, and patients in the combination therapy cohort (n=60) received up to 8 doses of RP2 and nivolumab starting with the second dose or third dose of RP2. RP2 was administered by direct intratumoral injection into subcutaneous lesions or into deep visceral lesions using imaging guidance. There were no significant differences between the two patient groups. 55.3% of the total patient population had an ECOG PS of 0, the main cancer types represented were uveal melanoma (20.0%), colorectal cancer (16.5%), head and neck cancer (15.3%), pancreatic cancer (14.1%), cutaneous melanoma (12.9%), and sarcoma (8.2%). Patients had a median of 2 prior lines of systemic therapy, ranging from 1 to 8, and 38.8% of patients had received more than 2 prior lines. 42.4% of patients had prior exposure to immune checkpoint inhibitors (ICI). RP2 monotherapy and combination therapy was well tolerated. The treatment-related adverse event (TRAE) profile of RP2 was consistent with systemic immune activation, and the most common TRAE was pyrexia, which occurred in 49.4% of patients. There was no grade 4 or grade 5 TRAEs, and the combination of RP2 with nivolumab was not significantly different from the expected safety profile of nivolumab alone. Patients received a median of 5 injections of RP2, ranging from 1 to 16, and 68.2% of injections were deep or visceral. The objective response rate (ORR) was 19.0% in the monotherapy cohort and 19.1% in the combination therapy cohort. One patient with esophagogastric adenocarcinoma in the monotherapy cohort achieved a complete response, and partial responses occurred in one patient with uveal melanoma, one patient with chordoma, and one patient with mucoepidermoid carcinoma. The median duration of response in the monotherapy cohort was not reached, but responses in this cohort lasted from 11 to 27+ months. Tumor regression was observed in both injected and non-injected lesions in patients from both cohorts. Responses were observed in the total patient population in a variety of tumors including uveal melanoma (n=5), cutaneous melanoma (n=4), head and neck cancer (n=1), and sarcoma (n=1), and no responses were observed in colorectal tumors or pancreatic tumors. RP2 monotherapy was associated with remodeling of the tumor microenvironment, with increased infiltration of CD8+ T cells, upregulation of PD-L1 expression, and activation of inflammatory pathways. Tumors exhibited increased gene expression signatures related to T cell cytotoxicity, antigen presentation, and antigen presentation, demonstrating the mechanism of action of anti-CTLA-4 expressed by RP2. Peripheral blood from responding and non-responding patients treated with RP2 monotherapy exhibited increased expansion of HSV-1-specific and MAGE-specific T cell receptor (TCR) clones from baseline to Day 43 of treatment. RP2 monotherapy and RP2 combined with nivolumab were well tolerated and produced durable responses in a heavily pre-treated patient population with diverse tumors. Investigation of RP2 for uveal melanoma is ongoing (NCT06581406), and additional studies of RP2 for other solid tumors are being planned.
Biomarker-driven TIL therapy for immunotherapy-resistant metastatic cancers
2505. Phase 2 basket trial of precision TIL therapy for immunotherapy-resistant advanced cancers
Udai Kammula (University of Pittsburgh Medical Center Hillman Cancer Center, Pittsburgh, PA, USA) reported initial results from a pilot phase 2 tumor infiltrating lymphocyte (TIL) therapy using TILScore as a precision biomarker enabling TIL therapy for immunotherapy-resistant cancers (NCT03935893). This was a Phase-2 Investigator-initiated basket trial of biomarker-guided TIL Therapy. TILScore is a pan-cancer transcriptomic biomarker that identifies metastases enriched with clinically active TILs, potentially increasing the feasibility and scalability of TIL manufacturing across solid tumors. This study builds on previous observations that some lesions from cancers that are typically immunotherapy-resistant do contain tumor-reactive TILs. In a past study of TIL therapy for metastatic uveal melanoma, a disease that is typically immunotherapy-resistant, infusion with tumor-reactive TILs resulted in an objective response rate of 43%, compared to an ORR of 0% with non-reactive TILs (n=52, p<0.01). TILScore is a histology-agnostic transcriptomic biomarker for predicting TIL potency before surgical procurement of tumors. TILScore is based on 265 immune-related transcripts derived from a preoperative tumor biopsy, and the model was trained on metastatic tumor transcriptomes from 14 solid tumors. A TILScore greater than 0.25 was identified as a threshold for predicting tumor reactivity. In an analysis of 490 metastases, over 50% of treatment-resistant metastases contained potentially tumor-reactive TILs (TILScore > 0.25). The cancer with the highest median TILScore was cutaneous melanoma, but metastases from historically immunotherapy-resistant cancers such as pancreatic ductal adenocarcinoma (PDAC), mesothelioma, and colorectal cancer also exhibited TILScores over 0.25. The single-center basket trial included 10 disease baskets that were refractory to front-line therapy. Prospective patients were screened for TILScore via tumor core biopsy. Patients with TILScore of 0.25 or higher were eligible for tumor resection and biomarker-guided TIL manufacturing. Eligible patients underwent lymphodepletion, followed by TIL infusion and high-dose supportive IL-2 therapy. 19 patients have been treated, representing PDAC (n=7), melanoma (n=4), sarcoma (n=3), squamous cell carcinoma (n=3), and peritoneal mesothelioma (n=2). Patients were heavily pre-treated, with a median of 6 prior therapies, ranging from 1 to 14. 47.4% of patients had prior exposure to immune checkpoint inhibitors, and TILs were harvested from liver (n=7), soft tissue (n=6), and lung (n=6). 14 patients had a TILScore of 0.25 or above, and 5 patients, who were included in the study for feasibility, had a TILScore below 0.25. Median tumor mutation burden was 2.1. Treatment-emergent adverse events (AE) were expected and consistent with non-myeloablative lymphodepletion. Grade 3 or 4 hematologic toxicities occurred in all patients, as expected, no unexpected AEs or grade 5 AEs were observed. The confirmed ORR was 33%, with 3 complete responses and 3 partial responses. All responses occurred in patients with high TILScores (>= 0.25), and the ORR was 46% in this patient population. Responses occurred across histological subtypes, including PDAC and peritoneal mesothelioma. Median duration of response was 7.2 months, with responses ranging from 3 months to 41+ months. This novel study indicates that TILScore-guided selection of metastases for TIL manufacturing is feasible and consistently enables generation of clinically active TILs across a variety of immunotherapy-resistant tumor types. Based on these results, the PDAC and mesothelioma cohorts met the pre-specified criteria for expansion, and the basket trial is ongoing and investigating TILScore-guided therapy for larger, histology-specific cohorts.
Early efficacy and safety results from inMMyCAR, a first-in-human study of BCMA-targeting CAR T cell therapy for relapsed and refractory multiple myeloma
7509. Updated results from inMMyCAR, the ongoing first-in-human phase 1 study of KLN-1010 in patients with relapsed and refractory multiple myeloma (RRMM)
P. Joy Ho (Royal Prince Alfred Hospital and University of Sydney, Camperdown, NSW, Australia) presented updated results from inMMyCAR, a first-in-human phase 1 study of in vivo CAR T therapy (KLN-1010) for relapsed and refractory multiple myeloma (RRMM). BCMA-targeting CARs were delivered in vivo with KLN-1010, a modified lentiviral vector that features a modified viral envelope and a CD3 tropism molecule for T cell targeting. Results were reported for the first 18 patients treated with KLN-1010. Patients had received a median of 3.5 prior lines of therapy, ranging from 2 to 7. 56% of patients were triple-class refractory, and patients had no prior exposure to BCMA-targeting agents. Patients did not undergo lymphodepletion prior to infusion, patients were treated across 3 dose levels, and the median time from consent to infusion was 13 days. At a median follow-up of 2.8 months, the objective response rate (ORR) was 100%. The first 14 patients who were treated are all minimal residual disease (MRD)-negative at 10^-5 or deeper, and MRD-negative bone marrow responses have sustained through six months. Some early partial responses have deepened to complete responses over time, and among the 6 patients with 4 or more months of follow-up, 4 have achieved a stringent complete response, and 2 have achieved a complete response. The earliest treated patients have sustained responses for 9 months or more. One patient has experienced regression after a complete response, and one patient has reverted to MRD-positivity below the limit of quantitation. All patients tested exhibit a decrease in soluble BCMA from baseline levels (n=9), and robust levels of CAR T expansion in bone marrow and blood (n=16). CAR T cells exhibited persistence in bone marrow and blood, with CAR T cells persisting in peripheral blood past 15 days, consistent with the potential for durable remissions. Toxicities were manageable and supportive of outpatient dosing. All patients experienced cytokine release syndrome, but all cases were grade 1 or 2 and manageable with standard care. 2 cases of immune effector cell-associated neurotoxicity syndrome (ICANS) occurred, and the single case of grade 3 ICANS was manageable and limited to 3 days. No delayed neurotoxicities were observed. No infusion-related reactions occurred after implementing premedication with dexamethasone. Cytopenias and infections of grade 3 or higher were limited. With its promising early clinical activity, potential for durable remissions, and favorable safety and tolerability profiles, in vivo CAR T therapy with KLN-1010 may potentially represent a novel off-the-shelf product that makes CAR T therapy more accessible for patients with RRMM.
To learn about a lipid nanoparticle-based system for generating in vivo CAR T therapies, see this paper from Fernández Bimbo et al in JITC.
Immune interception for oral premalignant lesions
10517. Intralesional PD-1 blockade for oral cancer prevention: First-in-class phase 1 trial
Moran Amit (The University of Texas MD Anderson Cancer Center, Houston, TX, USA) presented results from a phase 1, open-label, dose-escalation trial of intralesional nivolumab in patients with histologically confirmed oral epithelial dysplasia. Oral epithelial dysplasia represents a precancerous condition, and if oral premalignant lesions progress to cancer, surgery that can affect speech, swallowing, and oral competence, is the only standard of care. 29 patients with high-risk lesions or history of oral cancer were randomized to receive 10 mg (n=13) or 20 mg (n=16) intralesional nivolumab every three weeks for four cycles. 19 patients had a history of oral squamous cell carcinoma, and 10 did not. Treatment was well-tolerated: no dose-limiting toxicities occurred, 94% of adverse events were grade 1 or 2, and one patient in the 10 mg cohort discontinued treatment due to adverse events. Treatment was feasible, with high levels of patient adherence, despite a median one-way commute of 205 miles (ranging from 5 to 1,240 miles) for treatment and follow-up. Patient quality of life improved or remained stable. The clinical response rate with 85%, with responses occurring in patients receiving either dosage of nivolumab, and in patients with or without a prior history of oral cancer. Lesion area decreased by a median of 60%, and 21% of patients experienced a complete pathologic response. Among complete responders, 12-month disease-free survival was 75.8%. Spatial transcriptomics and multiplexed immunofluorescence indicated immune activation exclusively in treated lesions, including increased infiltration of CD4 and CD8 T cells, increased levels of CCR7+ activated dendritic cells, and formation of higher-order immune assemblies. These findings indicate that PD-1 blockade safely reprograms premalignant tissue without systemic toxicity and may allow some patients to avoid surgery and oral resections. A randomized, placebo-controlled phase 2 trial of intralesional nivolumab for high-risk oral premalignant lesions is currently enrolling patients, and this interventional approach may be investigated for other epithelial precancers.
Interim overall survival analysis of ivonescimab with chemotherapy versus tislelizumab with chemotherapy for previously untreated advanced squamous non-small cell lung cancer
LBA4. Ivonescimab plus chemotherapy versus tislelizumab plus chemotherapy in previously untreated advanced squamous non-small cell lung cancer: Overall survival results of the phase 3 HARMONi-6 trial
Shun Lu (Shanghai Chest Hospital, Jiao Tong University School of Medicine, Shanghai, Shanghai, China) presented results from the prespecified overall survival (OS) interim analysis of HARMONi-6, a randomized double-blind phase 3 study of ivonescimab plus chemotherapy versus tislelizumab plus chemotherapy for previously untreated advanced squamous non-small cell lung cancer (NSCLC). Ivonescimab is a first-in-class PD-1/VEGF dual target bispecific antibody that has been approved in China for two lung indications since 2024. A previous interim analysis (median follow-up 10.3 months) indicated that HARMONi-6 met its primary endpoint, with ivonescimab plus chemotherapy significantly improving progression-free survival (PFS) compared to tislelizumab plus chemotherapy (median PFS 11.1 months versus 6.9 months, p < 0.0001). Patients with pathologically confirmed stage IIIB-IV squamous NSCLC and no prior systemic therapy were randomized to receive up to 4 cycles ivonescimab with carboplatin and paclitaxel every three weeks, followed by ivonescimab every three weeks for up to 24 months (ivonescimab + chemo; n=266) or up to 4 cycles of cycles tislelizumab with carboplatin and paclitaxel every three weeks, followed by tislelizumab every three weeks for up to 24 months (tislelizumab + chemo; n=266). Treatment is ongoing in 60 patients in the ivonescimab + chemotherapy arm and 51 patients in the tislelizumab + chemotherapy arm. The two treatment arms were balanced with age, sex, ECOG as well as smoking history. 92.1% of patients in the ivonescimab + chemo arm and 91.5% of patients in the tislelizumab + chemo arm had stage IV disease. 39.5% of patients in each arm were PD-L1-negative (TPS < 1%), and 60.5% of patients in each arm were PD-L1-positive (TPS >= 1%). At a median follow-up of 21.36 months, ivonescimab + chemo significantly improved OS compared to tislelizumab + chemo, with median OS of 27.89 months and 23.69 months, respectively (HR 0.66; 0.50-0.87; p=0.0017). In all subgroups analyzed, an OS benefit favored ivonescimab + chemo. Ivonescimab + chemo improved across PD-L1 expression subgroups (HR 0.64 in the PD-L1 TPS < 1% subgroup; HR 0.68 in PD-L1 TPS >= 1%; HR 0.67 in PD-L1 TPS 1-49%; HR 0.64 in PD-L1 TPS >= 50%). 35.7% of patients in the ivonescimab + chemo arm and 36.5% of patients in the tislelizumab + chemo arm received subsequent anti-tumor systemic therapy, and 13.9% and 19.2% of patients, respectively, received subsequent immunotherapy. Ivonescimab + chemo exhibited a manageable safety profile consistent with prior reports and like the tislelizumab + chemo arm. 69.2% and 58.9% of patients, respectively, experienced grade 3 or higher treatment-related adverse events (TRAE). 5.3% and 4.5% of patients, respectively, experienced TRAEs that led to ivonescimab or tislelizumab discontinuation, and 14% of patients in each arm experienced an immune-related AE of grade 3 or higher. As expected, possible VEGF-related AEs occurred more frequently in the ivonescimab + chemo arm. 24.8% of patients in the ivonescimab + chemo arm and 12.1% of patients in the tislelizumab + chemo arm experienced hemorrhage, and 2.6% and 0.8% of patients, respectively, experienced hemorrhage of grade 3 or higher. Apart from alopecia, the most common TRAE were hematologic events, including anemia, neutropenia, and decreased white blood cell counts. Although longer follow-up is needed to validate the significance of the OS benefits associated with ivonescimab and chemotherapy, these data support ivonescimab with chemotherapy as an effective first-line treatment option for Chinese patients with advanced squamous NSCLC. A global randomized phase 3 trial (HARMONi-3; NCT05899608) of ivonescimab with chemotherapy versus pembrolizumab with chemotherapy for first-line treatment of metastatic NSCLC is underway.
To learn more about dual blockade of PD-(L)1 and VEGF for lung cancer, read this clinical study of atezolizumab (anti-PD-L1) and bevacizumab (anti-VEGF) for metastatic NSCLC by Ghiringhelli et al in JITC.
The effects of neoadjuvant durvalumab on tumor microenvironment features and their association with event-free survival in patients with resectable non-small cell lung cancer
8015. Impact of neoadjuvant durvalumab on tumor microenvironment features and their association with event-free survival in patients with resectable NSCLC from the phase 3 AEGEAN trial
John V. Heymach (The University of Texas MD Anderson Cancer Center, Houston, TX, USA) presented exploratory transcriptomic analyses of the tumor microenvironment (TME) in tumor samples collected at baseline and surgery to investigate the impact of neoadjuvant durvalumab on the TME and the association TME features with event-free survival (EFS). Patient samples were from the phase 3 AEGEAN trial, which previously showed perioperative durvalumab and neoadjuvant chemotherapy significantly improved EFS and pathological complete response compared to neoadjuvant chemotherapy alone in patients with resectable non-small cell lung cancer (R-NSCLC). Previous studies indicate that tumors with higher levels of immune cell infiltration at baseline (“hot tumors”) are more likely to respond to immune checkpoint blockade compared to immunologically “cold tumors.” Transcriptomic data were available for 366 samples from 292 patients across both arms of AEGEAN, comprising the RNA-sequence BEP. Baseline samples were analyzed from 257 patients, surgery samples were analyzed from 109 patients, and 74 paired samples (baseline + surgery) were analyzed. Baseline samples exhibited three distinct TME phenotypic clusters: Cluster 1 (C1; 24.9% of patients across both arms) was an immune desert consisting predominantly of proliferating tumor cells. Cluster 2 (C2; 39.3%) was immune suppressed, characterized by high levels of suppressive myeloid cells, epithelial-mesenchymal transition cells, angiogenic cells, and fibroblasts. Cluster 3 (C3; 35.8%) was immune activated, characterized by high levels of effector T cells. Further analyses found that TME clusters were consistently associated with specific histology and disease-related subgroups. Patients with C1 (immune desert) tumors were more likely to have a squamous histology, stage III disease, and N2 nodal status. C2 (immune suppressed) tumors tended to be associated with stage II disease, N0 nodal status and PD-L1 negativity. C3 (immune activated)tumors were more likely to have non-squamous histology, stage III disease, N2 nodal status, and high levels of PD-L1 expression. Addition of durvalumab improved EFS outcomes for all baseline clusters, especially C1 (immune desert; HR 0.43; 0.19-0.94) and C3 (; HR 0.41; 0.20-0.81) tumors, with the least benefit for C2 (immune suppressed) tumors (HR 0.90; 0.50-1.63). Durvalumab was also associated with higher pathologic complete response rates for all baseline clusters. The same clusters were observed in surgery samples (C1 32.1%; C2 33.9%; C3 33.9%), and patients with C3 (immune activated) tumors at surgery had the greatest EFS benefit with durvalumab but not with placebo. Patients with C1 (immune desert) tumors were at the highest risk of disease recurrence or death in either treatment arm. Although the sample size was limited, durvalumab was associated with strong EFS benefits in C3 (immune activated) tumors at surgery (HR 0.16; 0.03-0.79), compared to C1 tumors (immune desert; HR 1.0; 0.41-2.43) and C2 tumors (immune suppressed; HR 1.22; 0.38-3.99) at surgery. No patients with C1 (immune desert) tumors at surgery achieved a pCR. Paired samples from the durvalumab arm (n=32) exhibited a decrease in C1 (immune desert) tumors between baseline and surgery (28.1% to 21.9%) and an increase in C3 (immune activated) tumors between baseline and surgery (31.3% to 43.8%), suggesting neoadjuvant durvalumab acts to sustain and promote an immune activated TME. In contrast, paired tumors from the placebo arm (n=42) exhibited an increase in C1 (immune desert) tumors between baseline and surgery (33.3% to 52.4%). Although more analyses are needed to confirm and validate these findings, this analysis suggests that neoadjuvant durvalumab and chemotherapy, but not chemotherapy alone, can remodel the TME to a more immune activated state, which is associated with improved EFS in patients with R NSCLC.