Blood Test Maps Tumour Cell Neighbourhoods to Predict Immunotherapy Response

A New Window Into Tumour Biology

Researchers at Stanford Medicine and the Mayo Clinic have unveiled the first noninvasive blood test capable of mapping the cellular 'neighbourhoods' that surround a tumour, a breakthrough that could transform how oncologists predict immunotherapy outcomes. Published in Nature on 6 May, the study identifies 9 recurring multicellular ecosystems, dubbed spatial ecotypes, that recur across cancer types and can be detected from a routine blood draw using artificial intelligence.

How the Science Works

The team, led by Aaron Newman of Stanford and Aadel Chaudhuri of the Mayo Clinic, analysed over 10 million single-cell and spatial transcriptomes from 132 tumour specimens spanning 10 cancer types. Each spatial ecotype represents a microscopic neighbourhood, roughly the diameter of a human hair, defined by specific blends of immune, stromal, and cancer cells with distinctive gene expression patterns. The researchers then built Liquid EcoTyper, an AI framework that reconstructs these spatial patterns from methylation signals on circulating cell-free DNA in blood plasma, eliminating the need for invasive tissue biopsies.

Outperforming Existing Biomarkers

In clinical validation involving more than 1,300 patients with melanoma, lung, bladder, and gastric cancers, specific spatial ecotypes were strongly linked to immunotherapy response and survival. In pre-treatment plasma from nearly 100 melanoma patients, the liquid ecotype signals achieved a mean cross-cancer area under the curve of 0.87 for the response-associated ecotype SE7, far surpassing tumour mutational burden, PD-L1 expression, and circulating tumour DNA as predictive markers. Across solid tumours, liquid spatial ecotype signals were more significantly associated with overall survival than either TMB or PD-L1.

From Lab to Clinic

LiquidCell Dx, a diagnostics company co-founded with Newman's involvement, has announced it is developing LiquidTME, a clinical assay built on these findings to guide treatment decisions across cancer types. Newman envisions a future where clinicians use the blood test both to choose initial treatments and to track how the tumour microenvironment evolves over time, offering what he calls a more holistic and powerful method than anything currently available.