The Prognostic Value of Proclarix in Prostate Cancer Patients Under Active Surveillance: Predicting Transition to Active Treatment and Disease Progression in a Danish Cohort

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Recommendation:

• Tone down causal language.
• Emphasize hypothesis-generating nature more consistently.

There is lack of Multivariable Analysis. This is the biggest scientific limitation.

• Authors have only used:
• Kaplan-Meier
• Cox models (unclear if adjusted)
• There is no multivariable adjustment for:
• PSA
• Age
• Grade Group
• MRI findings

Without this, independent prognostic value cannot be claimed.

Required:

• Add multivariable Cox regression
• Show:
• Adjusted HRs
• Whether Proclarix adds value beyond clinical variables.

There are multiple small errors like :

• “Active active management” → ❌
• “assisst” → ❌
• “dicision-making” → ❌
• “risk-stratifications nomograms” → ❌

      Use consistent terms:

• “transition to AM” vs “switch to AM”
•  Standardize abbreviations early

 

Specify:

• Whether Cox models were adjusted or unadjusted
• Handling of missing data
• Whether proportional hazards assumption was tested
• Figures are referenced but no mention of:
• Number at risk tables
• Confidence intervals in KM curves.
• Improve figure reporting.

There are issues with table 1 :

• “NA 19 (14%)” → unclear what NA refers to
• Clinical staging categories are messy (cT1, cT1b, etc.)

 

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Reviewer Report

This manuscript evaluates the prognostic value of the Proclarix risk score in men with prostate cancer undergoing active surveillance, using a Danish cohort with long-term follow-up. The study addresses an important clinical question of how to better stratify patients for monitoring intensity and predict transition to active management or disease progression. The authors demonstrate that higher baseline Proclarix scores are significantly associated with both transition to active treatment and progression to clinically significant prostate cancer.

The manuscript has several strengths:

The study design is clinically relevant, focusing on a well-defined active surveillance population with real-world characteristics. The use of prospectively collected samples with retrospective blinded biomarker assessment strengthens the validity of the findings.

The follow-up duration (up to ~9 years) is a notable strength and allows meaningful evaluation of long-term outcomes. The statistical analyses, including Kaplan–Meier and Cox regression models, are appropriate and clearly presented.

Importantly, the results show a strong separation between risk groups, with high Proclarix scores associated with substantially increased risk of both treatment escalation and histological progression, suggesting potential clinical utility for patient stratification.

However, there are some limitations that can be addressed:

The retrospective nature of the analysis and the relatively modest sample size (particularly the number of progression events) limit the strength of the conclusions. In addition, the study would benefit from a clearer comparison with existing biomarkers and risk tools used in the active surveillance setting, to better contextualize the added value of Proclarix.

While the authors discuss this to some extent, a more quantitative or direct comparison would strengthen the manuscript. The choice of the 50% cut-off, although clinically intuitive, could also be further justified or explored (like sensitivity analyses).

Furthermore, while the clinical implications are promising, the conclusions should remain appropriately cautious given the need for prospective validation.

The manuscript is generally well written and structured, though minor improvements in clarity and tightening of the discussion would enhance readability. Some sections of the discussion could be streamlined to avoid repetition and maintain focus on the key findings and their clinical implications.

In conclusion, this is a well-conducted and clinically relevant study that provides encouraging evidence for the prognostic utility of the Proclarix test in the active surveillance setting. With minor revisions to strengthen contextualization, clarify methodological choices, and slightly temper conclusions, the manuscript would be suitable for publication.

 

Author Response

please see the attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This study addresses a clinically relevant question: non-invasive biomarkers that can help personalise active surveillance are clearly needed. The use of the PerPros biobank is a real strength and the blinded Proclarix measurements and the median follow-up of 6.3 years also strengthen the study. That said, several issues should be addressed before the manuscript is suitable for publication.

For me, the most important limitation is the absence of multivariable analysis as all Cox models appear to be univariable. As a result, it is not possible to determine whether Proclarix provides prognostic value beyond PSA density, clinical stage, or baseline Grade Group. These variables are already available at diagnosis and are part of routine clinical assessment. A multivariable model is therefore essential. Without it, the claim that Proclarix can guide clinical decision-making is not adequately supported.

As well, the 50% cut-off was developed and tested in the same cohort. This creates a clear risk of overfitting, which is a common concern in biomarker studies. The manuscript does not report any internal validation, such as bootstrap resampling or cross-validation. In addition, the high-risk group includes only 20 patients. Estimates such as “82% progressed to AM” are therefore based on very small numbers and may not be stable. The authors should report confidence intervals for all subgroup proportions.

The primary endpoint, transition from AS to AM, also needs better definition. The manuscript does not clearly separate these scenarios. E.g., a sensitivity analysis limited to pathologically confirmed progression would greatly strengthen the conclusions.

The manuscript is also missing standard biomarker performance metrics. It does not report AUC, sensitivity, specificity, or predictive values. These are expected in a biomarker evaluation study. Time-dependent ROC analysis would be especially useful in this setting. A direct comparison with PSA density within the same cohort would also help place the performance of Proclarix in context. The necessary data for PSA density analysis should already be available.

The proportional hazards assumption does not appear to have been tested? The Kaplan–Meier curves for csPCa progression seem to converge during later follow-up. This raises concern that the assumption may not hold. Schoenfeld residual testing should therefore be performed.

Several typographical errors also need correction. In the Simple Summary, “AS o AM” should be changed to “AS to AM.” “Assisst” should be corrected to “assist.” In the Abstract, “Dicision-making” should be changed to “decision-making.”

Author Response

please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors responded to all the comments nicely and the revised submitted manuscript now looks good.