DSC3-targeted immunotherapy shows promise in recurrent/metastatic HNSCC




Immunotherapy directed at desmocollin-3 (DSC3) safely improves outcomes in patients with recurrent or metastatic (R/M) head and neck squamous cell carcinoma (HNSCC) expressing DSC3, reports a study presented at ASCO 2026.
“DSC3 is a surface protein expressed by HNSCC,” according to the authors led by Dr Ranjan Das, Department of Medical Oncology, Sparsh Hospitals and Critical Care, Bhubaneswar, India.
“DSC3-directed immunotherapy (CADI-05) is useful in DSC3-expressing cancer like squamous non-small cell lung cancer, bladder cancer, and melanoma,” they added.
This single-arm study enrolled 20 patients with R/M HNSCC expressing DSC3 (mean age 55 years, 17 males) between May 2024 and July 2025. They received CADI-05 0.1 ml every 4 weeks and were followed up until disease progression or death. Das and his team also evaluated tumour infiltrating lymphocytes (TIL), PD-L1, and P16.
The primary sites of HNSCC were tongue (n=9) and buccal mucosa (n=9). All patients received two lines of therapy (mean 2.55). Of these, 15 had progressed on prior therapy. [ASCO 2026, abstract 2503]
Baseline parameters were as follows: mean DSC3 tumour proportion score (TPS) 41 percent (95 percent confidence interval [CI], 32.67‒49.33), mean PD-L1 combined positive score (CPS) 27 (95 percent CI, 10.61‒33.39), mean TIL 17 percent (95 percent CI, 10.43‒23.57).
Additional therapy received included EGFR tyrosine kinase inhibitor (n=15), EGFR monoclonal antibody (n=2), oral metronomic therapy (n=1), and none (n=2).
Outcomes
Of the patients with R/M HNSCC, nine (45 percent) achieved disease control, with overall response rate (ORR) of 35 percent (7/20: two complete and five partial responses).
The mean progression-free (PFS) and overall survival (OS) over a median follow-up of 155 days were 165 and 197 days, respectively. The mean duration of response was 280 days, with eight patients surviving at the last follow-up.
“This is significantly better than achieved with current therapies,” said Das, noting that R/M HNSCC are “aggressive tumours.”
In pretreated R/M HNSCC, the observed response rates were 13.0 percent with nivolumab (CheckMate 141) and 14.6 percent with pembrolizumab (KEYNOTE-040), with median PFS of 2.0 months and 2.1 months, respectively, according to Das and colleagues.
Notably, DSC3 expression was higher among responders than non-responders (mean 50 percent vs 29 percent; p=0.002537). ORR was higher in patients with HNSCC stage IVA than those with stage IVC (57 percent vs 20 percent).
Among patients with disease control, DSC3 TPS was higher than PD-L1 CPS. TIL and PD-L1 levels showed no significant association with response.
With regard to safety, the only side effect recorded was injection site ulcer, which occurred in eight patients (40 percent), including six (75 percent) of the responders. Systemic adverse effects were not observed.
“DSC3-targeted immunotherapy was safe and improved outcomes in pretreated R/M HNSCC expressing DSC3,” Das and his team said.
The study was limited by the lack of a control arm, heterogeneous group in terms of line of therapy, the less frequent use of DSC3-specific therapy (once a month), and the companion therapy variable, according to Das.