Comment on ‘Novel MRI‐US fusion with advanced annotation in focal cryoablation for prostate cancer: medium‐term outcomes and learning curve insights’

We read with great interest the single‐centre prospective series by Marquis et al. [1] describing MRI‐US fusion guided focal cryoablation (FC) enabled by a three‐dimensional Advanced Annotation Platform (AAP). The authors report encouraging medium‐term outcomes—2‐year failure‐free survival (FFS) of 96%, in‐field recurrence of 6% and out‐of‐field recurrence (OFR) of 12% at the 1‐year protocol biopsy—together with favourable functional recovery, only Grade 1–2 complications, and minimal differences between an expert and a beginner surgeon. This work highlights the promise of digital planning and guidance to standardise FC workflows while maintaining safety and efficacy.

We wish to offer several suggestions that, in our view, could further strengthen the clinical message and facilitate broader adoption. First, clarification of endpoint definitions using sensitivity analyses would help align oncological control with patient‐level treatment burden [2]. Because the primary FFS did not count a second focal re‐ablation as failure, readers would benefit from a parallel presentation in which any tumour‐directed re‐intervention constitutes failure, shown alongside the current primary definition. Displaying both Kaplan–Meier curves upfront would provide complementary perspectives that map more directly onto real‐world decision‐making regarding when to move from repeat focal therapy to radical treatment.

Second, the OFR signal merits a more standardised surveillance pathway. The protocolised 1‐year systematic biopsy is a clear strength; however, beyond this point follow‐up was largely for‐cause. Given the observed OFR rate, a scheduled assessment window (e.g., multiparametric MRI with or without biopsy at 18–24 months) could reduce heterogeneity and missed events [3]. A pragmatic risk score that integrates Prostate Imaging‐Reporting and Data System (PI‐RADS) category, PSA density, lesion volume/site and grade might further triage follow‐up intensity and help tailor when a patient should undergo confirmatory biopsy vs continued imaging.

Third, one of AAP's unique values is its digital trace across planning and execution. We encourage routine reporting of objective image‐to‐treatment quality assurance (QA) metrics—such as registration root‐mean‐square error (mm), the proportion of target volume encompassed by the planned 100% isotherm (%), needle tip/trajectory deviation (mm/°), and minimum distance to the urethral sphincter and rectum (mm)—and correlating these with IFR/OFR and complications [4]. This transparency would help distinguish biological failure from modifiable technical variance and would provide centres with concrete benchmarks during programme roll‐out.

Finally, extending the learning curve analysis from process to effect could add practical guidance. The clearest learning signal related to efficiency (ablation time) with smaller effects for probe number and OFR, and none for in‐field recurrence (IFR) or complications. Segmenting the series (piecewise regression) or adopting a hierarchical Bayesian framework—with the above QA metrics entered as covariates—could test for thresholds (‘how many cases until coverage is consistently optimal?’) and link incremental workflow gains to clinically meaningful outcomes.

In summary, Marquis et al. [1] provide important prospective evidence that AAP‐enabled MRI‐US fusion can deliver precise and safe FC with limited dependence on prior focal‐therapy experience. Clarifying failure definitions, formalising OFR‐oriented surveillance, and embedding objective QA and outcome‐level learning analyses would, we believe, make the findings even more actionable for centres seeking to implement FC programmes.

Disclosure of Interests

The authors declare no conflicts of interest.