The combination of pembrolizumab, an immune checkpoint inhibitor, with vorinostat, a drug that modifies gene expression, achieved response rates of up to 39 per cent in some cancer types during a phase 2 clinical trial. The findings suggest that priming the immune system with epigenetic modifications could overcome resistance to immunotherapy alone.
Squamous cell carcinomas affect various parts of the body including the cervix, anus, head and neck, and genital regions. While immunotherapy has revolutionized cancer treatment, only a minority of patients - typically 15 to 24 per cent - respond when these drugs are used alone.
"The combination strategy addresses a critical challenge in cancer immunotherapy," says Christophe Le Tourneau at Institut Curie in Paris, who led the international study. "By using vorinostat to modify the epigenetic landscape, we're essentially making tumors more visible to the immune system."
The PEVOsq trial enrolled 112 patients with advanced cancers that had returned or spread despite previous treatments. The researchers used a "basket trial" design, simultaneously testing the drug combination across multiple cancer types rather than focusing on a single disease.
The results varied significantly by cancer type. Cervical cancer patients showed the highest response rate at 39 per cent, followed by anal cancer at 31 per cent. However, the combination was less effective for head and neck cancers (19 per cent) and penile cancer (18 per cent).
"What's particularly intriguing is how the same drug combination can have such different effects depending on the cancer's origin," says Le Tourneau. This variability likely reflects the distinct molecular characteristics of squamous cell carcinomas from different body sites.
The treatment wasn't without challenges. Two-thirds of patients required dose reductions or treatment interruptions due to side effects from vorinostat, including nausea, fatigue, and blood abnormalities. Despite these issues, the side effects were generally manageable and no treatment-related deaths occurred.
Beyond testing the drug combination's effectiveness, the researchers conducted extensive genetic analyses to identify biomarkers that could predict which patients would benefit most. They discovered that patients with mutations in genes like B2M, RAD51, and NOTCH1 showed longer progression-free survival.
Patients with high tumor mutational burden - meaning their cancers had accumulated many genetic changes - were particularly likely to respond, with 58 per cent showing tumor shrinkage compared to just 20 per cent in those with fewer mutations.
HPV status also emerged as an important factor. Patients whose cancers were linked to human papillomavirus infection had better response rates (34 per cent) compared to HPV-negative cancers (16 per cent).
The findings open new avenues for combination therapies in cancer treatment. Vorinostat works by inhibiting enzymes called histone deacetylases, which can alter how genes are expressed without changing the DNA sequence itself. This epigenetic reprogramming appears to enhance the immune system's ability to recognize and attack cancer cells.
"We're moving beyond the one-size-fits-all approach to cancer treatment," notes Le Tourneau. "Understanding why certain cancers respond better to this combination could help us design more personalized treatment strategies."
The research team is now planning larger studies to validate these findings and explore whether the genetic biomarkers they identified can reliably guide treatment decisions. They're also investigating ways to reduce vorinostat-related side effects while maintaining the drug's cancer-fighting benefits.
For patients with advanced squamous cell carcinomas, particularly those with cervical or anal cancers, this combination therapy could offer a new treatment option where few currently exist. As our understanding of the complex interplay between epigenetics and immunotherapy grows, similar combination approaches may prove effective for other hard-to-treat cancers.
Journal reference: Nature Cancer DOI: 10.1038/s43018-025-01004-2