Design: Two nested case-control studies within national registries
Population: Primary cutaneous squamous cell carcinomas from 19,120 patients; 195 and 250 matched pairs
Primary outcome: Association of refined histological variables with metastasis
Effect: No refined variable significant in the population-based set; solar elastosis OR 0.20–0.28 in the risk-matched set

Using Dutch national cancer and pathology registries covering more than 19,000 patients followed for over ten years, two nested case-control studies matched primary tumours that later metastasised with ones that did not. One set was matched on follow-up time (195 pairs); the other on a calculated metastatic risk from conventional factors (250 pairs). Pathologists re-reviewed all primaries for morphological subtype, solar elastosis, peritumoral infiltrate, tumour budding and mitoses.

In the population-based set, none of these refined features was independently associated with metastasis. In the risk-matched set, moderate and extensive solar elastosis were associated with lower odds of metastasis, an unexpected finding that needs confirming.

For practice, this supports reporting and acting on the established features, such as depth of invasion, differentiation, perineural invasion and tumour size, rather than adding extra histological scoring to decide on staging or follow-up. It was published in September 2026.

Base squamous cell carcinoma risk assessment on established features: size, depth, differentiation and perineural invasion.
Tumour budding, peritumoral infiltrate and subtype did not add independent prognostic value here.
Treat the lower metastasis risk with solar elastosis as a hypothesis, not a reason to relax follow-up.
Use a recognised staging system to decide on nodal imaging and follow-up intensity.

Why it matters

It saves pathology effort that does not change a patient's risk estimate.

The statistics, in plain English

Odds ratios below 1 for solar elastosis (0.20 and 0.28) suggest lower odds of metastasis, but the intervals are wide (up to 0.87), reflecting small numbers. A nested case-control design compares cases with matched controls drawn from the same cohort, which is efficient but depends on good matching.