Ultrasound localisation microscopy tracks testicular microvascular adaptations to endocrine function in male infertility

A super-resolution ultrasound scan detected testicular treatment response before blood hormone tests did

Journal: EBioMedicine | Published: 2026-06-18 | Type: Journal Article | PMID: 42314538 Authors: Papanikolaou N, Yan J, Huang B, Davies R, et al. (Imperial College London and Imperial College Healthcare NHS Trust) Funding/COI: MRC, NIHR Biomedical Research Centre Funding Scheme, NIHR/Imperial Clinical Research Facility, Diabetes UK, BBSRC, and the Imperial Private Healthcare Clinical Research Fellowship Scheme. Co-author Mengxing Tang sits on the Scientific Advisory Board of Verasonics and holds minority shares in Sonalis Imaging and Cardioacc, all ultrasound imaging companies; declared unrelated to this work.

Summary

Researchers used ultrasound localisation microscopy (ULM), a technique that maps blood vessels at a resolution finer than standard ultrasound allows, to image testicular microvasculature in men with hypogonadotrophic hypogonadism (HH) versus healthy controls. Vessel density, diameter, and tortuosity differed between groups and tracked with testosterone and inhibin B levels, and in a small longitudinal cohort, ULM changes appeared before hormone or testicular volume changes during fertility treatment.

Claims

• ULM vessel density (p < 0.01), diameter (p = 0.01), and tortuosity (p < 0.01) discriminated HH (n = 9) from controls (n = 10)
• Vessel density and diameter correlated with testosterone (r = 0.53-0.69) and inhibin B (r = -0.61 to 0.65, p < 0.05 to p < 0.001) across studies
• Vessel density, diameter, area, and flow index were lower in azoospermic men regardless of treatment type (p < 0.01)
• Across three HH treatment subgroups, TRT (n = 11), gonadotrophins (n = 9), and untreated (n = 12), ULM measures still separated groups by density (p < 0.001) and diameter (p < 0.05)
• In a 12-month longitudinal cohort (n = 7) starting gonadotrophin therapy, ULM parameters rose (p < 0.05) and reportedly picked up testicular activation before testicular volume or inhibin B did
• A parallel rodent pubertal-blockade model (n = 5) showed microvascular remodeling tied to testis growth

Study Quality

This is a prospective case-control design run across three linked human cohorts plus a small animal model, which is a reasonable way to build both cross-sectional and longitudinal evidence for a new imaging biomarker. The statistics used (t-tests, Mann-Whitney U) are appropriate for the data types described, and the correlations with testosterone and inhibin B give some biological plausibility to the ULM signal.

The problem is scale. Every subgroup is small: 9 to 12 men per arm, 7 men followed longitudinally, 5 rodents. With that many outcome measures (density, diameter, tortuosity, area, flow index, velocity) tested against multiple hormones across multiple substudies, the paper does not describe a correction for multiple comparisons, so some of the "significant" associations could be noise. There is also no mention of blinding for the researchers scoring ULM images against clinical group.

Red Flags

• Sample sizes are small throughout: as few as 7-12 men per group, 5 rodents in the mechanistic arm
• No stated correction for multiple statistical comparisons across many correlated outcome variables
• No described blinding of image analysts to case-control or treatment status
• Single center (Imperial College Healthcare NHS Trust), which limits generalizability
• A co-author has financial ties (advisory board seat, equity) to companies commercializing ultrasound imaging technology similar to what is being validated here, even though declared unrelated
• HH testes are smaller by volume than controls (7.5 vs 14.9 mL, p = 0.01), so some vessel density differences could partly reflect organ size rather than a distinct microvascular signature

Strengths

• Technical validation of the resolution claim: 90.3 um measured via Fourier ring correlation versus the 128.3 um classical ultrasound diffraction limit
• Multiple complementary study designs (cross-sectional case-control, treatment subgroup comparison, longitudinal follow-up, animal model) pointing the same direction
• Use of a standard, widely available clinical ultrasound scanner rather than bespoke research-only hardware
• Correlation with established, clinically used biomarkers (testosterone, inhibin B) rather than an isolated imaging metric
• Ethically approved prospective protocol with informed consent and accredited hormone and semen analysis labs

Verdict

This is a genuinely novel piece of engineering, applying super-resolution vessel mapping to an organ that has had few new diagnostic tools in decades, and the internal consistency across three human studies and an animal model is more than you'd get from a single small trial. But the samples are tiny, the multiple-comparisons problem is real, and a co-author's financial stake in ultrasound imaging companies deserves attention even if declared unrelated. Read this as an interesting proof-of-concept for a future diagnostic, not evidence that ULM is ready to replace hormone testing or testicular volume measurement in the clinic.