We appreciate Prof. R. Loffroy’s interest in our recent publication on prostatic artery embolization (PAE) using ethylene vinyl alcohol copolymer (EVOH) [1, 2]. In our opinion, several points warrant clarification.

The first point concerns the impact of EVOH on diagnostic imaging. The radiopacity of EVOH-based liquid agents is provided by micronized tantalum, a metal (Z = 73) whose deposition leads to persistent hyperdensities visible across all X-ray-based imaging modalities. In our series, some patients underwent post-PAE pelvic CT scans. Although no significant beam-hardening artifacts were observed (Fig. 1), we acknowledge that these hyperdensities may complicate interpretation in specific situations (e.g., detection of ureteral calculi). Similarly, EVOH may hinder cone-beam CT (CBCT) angiography analysis because of potential confusion with contrast medium, which may represent a limitation for repeat embolization after EVOH-PAE. In contrast, systematic pre- and 3-month post-procedural MRI demonstrated no relevant EVOH-related artifacts. This aligns with well-established neuroradiological data, where EVOH-embolized cerebral vascular malformations are routinely followed by MRI. In that setting, EVOH-filled vessels typically appear hypointense, without degradation of diagnostic quality from susceptibility artifacts, owing to the non-ferromagnetic nature of tantalum [3]. In our series, no such hypointensity was observed (Fig. 2), likely reflecting the small caliber of prostatic arteries and the limited amount of embolic material required (approximately 0.5 mL per hemiprostate). These findings suggest that EVOH preserves MRI diagnostic performance for prostate cancer detection—a prerequisite for any embolic agent used in this indication.

Fig. 1The alternative text for this image may have been generated using AI.

Pelvic CT scan in the venous phase after contrast injection, in the axial plane through the prostate, shows hyperdensities (500–5000 HU) along the periphery of the prostate related to prior EVOH-PAE. Note the absence of beam-hardening artifacts from EVOH deposits. EVOH, ethylene vinyl alcohol copolymer, PAE, prostatic artery embolization, HU, Hounsfield units

Fig. 2The alternative text for this image may have been generated using AI.

A, D: T2-weighted prostate MRI images in the axial plane through the mid-prostate (A) and median lobe (D), before PAE. B, E: Non-contrast-enhanced pelvic CBCT in the axial plane through the mid-prostate (B) and median lobe (E), at the end of PAE (final CBCT). C, F: T2-weighted prostate MRI images in the axial plane through the mid-prostate (C) and median lobe (F), 3 months after PAE. Note the absence of magnetic susceptibility artifacts related to EVOH deposits, which should not be confused with T2-hypointense infarcted areas (C, square). CBCT, cone-beam computed tomography, EVOH, ethylene vinyl alcohol copolymer, PAE, prostatic artery embolization

The second point relates to procedural duration and, consequently, radiation exposure. In our study, the median (interquartile range) fluoroscopy time and dose–area product were 25 min (19–32.5 min) and 25,600 µGy·m2 (19,000–36,200 µGy·m2), respectively. By comparison, Loffroy et al. reported corresponding values of 30.1 min (18–36.5 min) and 22,846 µGy·m2 (15,680–37,451 µGy·m2) for glue-PAE [4]. At first glance, the argument of a dosimetric penalty associated with EVOH does not appear self-evident. It should also be emphasized that the time-consuming “plug-and-push” technique was not used systematically in our study, but rather reserved for cases in which penetration achieved during initial injection—arguably the most critical step—was insufficient or could be improved. Although EVOH injection time was not specifically evaluated in our study, and glue delivery may indeed be faster, overall procedure duration remains mainly driven by the complexity of prostatic artery catheterization. Importantly, both liquid agents are inherently faster to deliver than microparticles. From a broader clinical perspective, the pursuit of a marginal reduction in radiation exposure should not outweigh the use of an embolic agent that may offer greater controllability [5].

The third point concerns pain related to dimethyl sulfoxide (DMSO), the solvent required to maintain the copolymer in a liquid state during injection. In our view, this represents the main drawback of EVOH-based agents, as their use requires a more elaborate analgesic protocol than local anesthesia alone. In our experience, a multimodal approach combining paracetamol, morphine, and intra-arterial lidocaine provided acceptable pain control. While some teams favor sedation or general anesthesia [6], this does not call into question the minimally invasive nature of PAE. Should interventional radiologists be restricted to subcutaneous lidocaine injection when performing outpatient procedures?

We concur with Prof. Loffroy that penetration and occlusion are distinct concepts: distal penetration does not necessarily guarantee effective vascular occlusion. Successful embolization primarily relies on complete luminal filling, a requirement that becomes even more critical as vessel caliber increases (e.g., proximal embolization of the main prostatic artery). However, the absence of a demonstrated association between penetration and clinical outcomes using the Glue Penetration Index Score cannot be taken as definitive evidence against such a relationship [7]. In this regard, it is important to emphasize that the Prostatic EVOH Penetration Score (PEPS) differs in several respects, notably in being based on cross-sectional imaging (CBCT), which we believe provides a more reliable assessment of penetration than a simple anteroposterior radiographic projection (Fig. 3). We therefore encourage the use of 3D imaging for this purpose, as well as for the reliable detection of non-target embolizations. While PEPS may appear reductive given the complexity of the phenomenon it seeks to capture, it conveys a clear message: clinical success requires intraprostatic penetration, and purely peripheral embolization may be insufficient. The contribution of other, potentially more subtle parameters—such as intraglandular distribution of the embolic agent, injected volume, homogeneity of vascular filling, or degree of network saturation—remains to be determined.

Fig. 3The alternative text for this image may have been generated using AI.

A Frontal pelvic radiograph at the end of PAE shows EVOH deposition over the prostate silhouette, extending to the midline (arrow). B Post-procedural non-contrast-enhanced pelvic CBCT with axial reconstruction through the mid-prostate shows EVOH within periprostatic vessels reaching the capsule on both sides, without intraprostatic penetration (PEPS 2|2) (1). The median vessel seen on the frontal radiograph corresponds to an anterior capsular, non-intraprostatic vessel. CBCT, cone-beam computed tomography, EVOH, ethylene vinyl alcohol copolymer, PAE, prostatic artery embolization, PEPS, Prostatic EVOH Penetration Score

Finally, the durability of clinical outcomes remains the key challenge of any technique intended to treat a chronic condition. Our results, while encouraging, are limited to a 3-month follow-up and must therefore be interpreted with appropriate caution, without dismissing the observed relationship between PEPS and early clinical success. It appears relevant to distinguish two dimensions of PAE effectiveness: the initial clinical response, partly driven by embolic penetration, and its durability, likely dependent on complete occlusion of the entire prostatic arterial supply.

In conclusion, our study was not intended to compete with glue, but to demonstrate the feasibility, safety, and short-term efficacy of EVOH-PAE, while exploring the concept of penetration, which is central to understanding liquid embolic agents regardless of their nature. We welcome the technical shift of PAE toward liquid agents, which may lead to improved and more durable clinical success through optimization of both penetration and vessel occlusion, two parameters that can be directly assessed during the procedure. The comparative safety and efficacy of different liquids can only be established through randomized head-to-head studies, which will hopefully be conducted in the near future.