Patients with advanced solid tumors often run out of treatment options when chemotherapy and immunotherapy fail. For a subset of these patients, a new Phase-1 study now shows a possible path forward. Researchers at the National Center for Tumor Diseases (NCT/UCC) at Dresden University Hospital presented data on IMA401, a bispecific T-cell engager from Tuebingen-based company Immatics, at the American Society of Clinical Oncology (ASCO) annual meeting on May 31, 2026, in Chicago. The study was simultaneously published in Nature Medicine. In head and neck tumors, the therapy achieved a confirmed response rate of 29 percent in patients who had previously received multiple other therapies that had all failed.
Why solid tumors are the harder challenge
Immunotherapy has revolutionized blood cancer treatment over the past decade. CAR-T cells, genetically modified immune cells from a patient's own body, achieve response rates exceeding 80 percent in certain B-cell leukemias and have saved patients for whom all other options were exhausted. For solid tumors, however, CAR-T cells have had little effect so far.
The reason lies in the tumor itself. Solid carcinomas develop a hostile microenvironment that rejects infiltrating immune cells or pushes them into an exhausted state. Moreover, it is difficult to find antigens that sit on the tumor but not on healthy tissue. Attacks on healthy tissue would be dangerous. IMA401 bypasses this problem by targeting an antigen that occurs rarely on healthy adult cells: the cancer-testis antigens MAGEA4 and MAGEA8.
What makes IMA401 different
IMA401 is neither a classical antibody nor a CAR-T cell, but a bispecific T-cell receptor engager (TCER). The molecule has two binding arms: one targets the MAGEA4/8 antigen on the cancer cell, the other binds the T-cell receptor on cytotoxic T lymphocytes. This brings the body's own killer cells into direct contact with cancer cells without the T cells having to be modified in the laboratory first, as is required for CAR-T therapy.
In the Phase-1 study design, researchers treated 61 patients with recurrent or refractory solid tumors across more than 15 tumor types. At the recommended Phase-2 dose, either alone or combined with the checkpoint inhibitor pembrolizumab, the head and neck tumor subgroup showed 4 out of 14 patients as responders. This corresponds to a confirmed response rate of 29 percent. The disease control rate was 64 percent. Median response duration was 8.8 months. Particularly striking: all responding patients showed tumor reductions of 60 to 100 percent. Tolerability was favorable according to the study data, with no cases of severe neurotoxicity (ICANS) observed.
By comparison: checkpoint inhibitors and the limits of current treatments
Checkpoint inhibitors like nivolumab and pembrolizumab were approved starting in 2014 and have changed treatment in many tumor types. In recurrent head and neck squamous cell carcinoma, pembrolizumab as monotherapy after platinum-based chemotherapy failure achieves response rates of approximately 15 to 16 percent, as the KEYNOTE-012 study demonstrated. This is meaningful for an advanced situation, but more than half of patients do not respond.
IMA401 shows double the response rates in a similar patient population, namely heavily pretreated patients with no further standard options. This is a Phase-1 result with a small patient cohort and must be confirmed in larger studies. But the signal magnitude is clinically relevant for an early study phase.
Another comparison point: in solid tumors, CAR-T cell clinical trials have largely failed to show convincing response rates. The combination of a TCER mechanism and an antigen rarely expressed on adult healthy tissue appears to partially overcome this problem.
Three hurdles until clinical use
The ASCO data are promising, but three steps separate IMA401 from regular approval. First, the Phase-2 study must reproduce results in larger patient groups. Phase-1 studies are designed for dose-finding and safety. The small patient numbers in individual tumor types do not yet allow reliable statistical statements. Immatics has announced opening a combination cohort pairing IMA401 and IMA402, another PRAME-directed TCER, for lung cancers. First data are expected in 2027.
Second, success depends on patient selection. IMA401 works through the MAGEA4/8 antigen, which not all tumor patients express. A reliable biomarker test is needed to identify which patients will benefit from the therapy before the drug is used more broadly.
Third, manufacturing bispecific proteins of this class is technically demanding and still expensive. Later market approval depends on whether the production process can be industrialized to the extent that the therapy becomes available in sufficient quantities and at sustainable costs. With these limitations, IMA401 remains one of the most interesting candidates in early clinical oncology, awaiting its next test.