Andor van den Hoven, MD, Describes a Particle-Fluid Dynamics Study for Radioembolization of Colorectal Liver Metastases

IO360: Please describe your facility, the makeup of the IO team, and the IO procedures done at your facility.

van den Hoven: I am an MD working as a PhD candidate at the Department of Radiology and Nuclear Medicine of the University Medical Center Utrecht in Utrecht, Netherlands. Currently, our interventional oncology team includes interventional radiologists Maurice van den Bosch, MD, PhD, and Rutger Bruijnen, MD; nuclear medicine physician Marnix Lam, MD, PhD; Maarten Smits, MD, PhD; Jip Prince, MD; and several interventional radiology fellows and other PhD candidates. Our main focus is on radioembolization for primary liver cancer and various types of liver metastases, and high-intensity focused ultrasound (HIFU) treatments for breast cancer, painful bone metastases, and renal cancer. In addition, we treat hepatocellular carcinoma (HCC) patients with radiofrequency ablation (RFA) and transarterial chemoembolization (TACE). These procedures are performed as routine clinical care and as part of our clinical trials, in close collaboration with nuclear medicine, diagnostic radiology, medical oncology, surgery, and radiotherapy specialties. 

We are also exploring new possibilities for interventional oncology such as image-guided, intratumoral injection of radioactive microspheres in patients with head and neck cancer.

IO360: What was the impetus for the study? Had you been noticing administration of microspheres that you weren’t happy with?

van den Hoven: Yes. The majority of patients that we treat with radioembolization in our center have colorectal liver metastases. Due to the hypovascular nature of these metastases, blood flows only marginally preferential toward tumorous tissue. Consequently, slight changes in catheter position and hemodynamics can have significant impact on the microsphere distribution. This is an important difference from other tumor types such as HCC or liver metastases from neuroendocrine tumors, uvea melanoma or cholangiocarcinoma, and poses two important problems during radioembolization. For one, the intrahepatic distribution of technetium macroaggregated albumin (MAA) as depicted on SPECT/CT is not a reliable predictor of the therapeutic microsphere distribution. Unfortunately, this also hinders the development of personalized dosimetry models. Second, tumor targeting is not always optimal in clinical practice. We have frequently observed lack of  Y90 activity in individual tumors on post-treatment Y90-PET/CT (click the figure for a clinical example). Using a more selective injection position would limit the influence of flow dynamics, but this is often not feasible without compromising tumor targeting in patients with an extensive disease burden, and the use of more than three injection positions is very impractical.

Based on our own experience with the novel Surefire Infusion System antireflux catheter (Surefire Medical) during radioembolization procedures, and promising data from preclinical and clinical studies demonstrating enhanced infusion efficiency and a down-stream pressure gradient when using the Surefire system, we came to believe that the antireflux catheter may have the potential to improve the predictive value of the intrahepatic scout dose distribution and tumor targeting during therapy. By design, the catheter outlet of the antireflux catheter is, in contrast to a standard microcatheter, fixed in the center of the vessel lumen. This guarantees the same cross-sectional catheter position during a scout procedure and a therapy procedure, which may improve the predictive value of the scout dose distribution. Furthermore, we noticed in clinical practice that the outflow pattern of contrast agent looked different on DSA when using an antireflux catheter, but we did not understand the physics behind this phenomenon and we did not know how this could affect tumor targeting during treatment.

Also, we realized that it is odd that so little is known about the interplay of catheter design, catheter positioning, and particle-fluid dynamics during radioembolization. So, we decided to team up with the excellent engineering partners Dr. Gregory Buckner and Shaphan Jernigan from the department of Mechanical and Aerospace Engineering at the North Carolina State University in Raleigh, North Carolina, and investigate catheter-related effects on particle-fluid dynamics in a surrogate model of the hepatic arterial vasculature with recreated hemodynamics.

IO360: How did you choose your study methods and the fluorescent holmium microspheres? Did you build the vascular model?

van den Hoven: The transparent vascular model and the setup for recreated hemodynamics had already been built and tested by Dr. Gregory Buckner and his team. In a previous study they used this model with an additional tumor model to compare tumor penetration between resin and glass Y90 microsphere administrations.¹ 

The model allowed us to perform multiple comparative experiments in a controlled environment. In total, we performed 7 experiments during which microspheres were administered with the antireflux catheter as well as the standard microcatheter. Several parameters such as microsphere type, catheter position and injection technique varied across experiments to address specific questions, but were kept constant between the comparative runs with the antireflux catheter and standard microcatheter. We chose to use fluorescent microspheres with a size and density profile comparable to resin Y90 microspheres during the first 3 experiments, because this allowed us to visualize the outflow pattern by video recording with an HD camera under UV-light illumination. We felt that this would enable us to perform a simple qualitative analysis of the particle outflow pattern for both catheter types, and effectively communicate our findings through images. In addition, we performed a quantitative analysis of the downstream branch distribution of microspheres by collecting the microspheres in separate filters connected to the target branches during all experiments. Experiments 4-7 were performed with holmium-166 (Ho166) microspheres to validate our findings with a microsphere type that we also use to treat patients in clinical practice.

We chose to align the longitudinal position of the antireflux and standard catheters, but we accepted random standard microcatheter tip deviation in the cross-sectional vessel plane, because catheter tip deviation likely also occurs in clinical practice without us noticing it on 2-dimensional DSA images, and it may impact downstream branch distribution. Furthermore, the effect of varying injection force (soft, average, and hard pressure) was investigated during manual syringe operation in the first experiment with the ARC and SMC. Automatic syringe operation (with a profile similar to clinical practice) was chosen as the preferred method for the rest of the experiments to prevent potential bias between the antireflux and standard catheter administrations.

IO360: Please share the most important results.

van den Hoven: We found that the antireflux and standard catheters have a strikingly different particle outflow pattern. The standard microcatheter administrations showed a small, fine-lined, and organized particle outflow pattern consistent with laminar flow, whereas the antireflux catheter administrations showed a broad, chaotic outflow pattern consistent with turbulent flow. The dynamically expandable tip of the antireflux catheter likely breaks up the laminar columns in the antegrade blood flow, thereby creating turbulence downstream of the catheter tip. 

Lateral catheter tip deviation of the standard microcatheter was noted in the majority of the experiments, and the downstream branch distribution was heavily skewed toward the side of deviation in these experiments. The downstream branch distribution was also dependent on injection force, with improving homogeneity of the distribution when increasing pressure. The antireflux catheter administrations were associated with a fixed centroluminal catheter position, and a more consistent and significantly more homogeneous downstream branch distribution, compared with the standard microcatheter.

IO360: What surprised you the most about the results?

van den Hoven: The striking difference in particle outflow pattern between the standard microcatheter and antireflux catheter administrations surprised me the most. I had not expected to see such a clear difference on the video recordings of the fluorescent microsphere administrations. What also surprised me was the effect that injection force had on the particle outflow pattern during the SMC administrations. But it makes sense. With soft to average injection force, the microspheres are carried by the laminar columns but are not able to cross streamlines laterally. In fluid mechanics, it is a known fact that fluid velocity affects the transition from laminar flow to turbulent flow.

IO360: What’s the important takeaway for IO clinicians about the results?

van den Hoven: I think that the most important takeaway is that catheter design, catheter positioning, and fluid dynamics do matter. We should try to further enhance our knowledge about the interplay between catheter and fluid-particle dynamics and find a way to use these principles of physics in order to improve patient care. The standard microcatheter, still our main tool in intra-arterial cancer treatments, may not suffice. Maybe different catheter types should be developed for different purposes.

Some interventional radiologists will already have experienced that the slightest adjustments in catheter position or injection force can significantly impact flow. Unfortunately, current technology does not allow us to generate an intuitive 3-dimensional image of the catheter position in the vessel lumen, and real-time monitoring of flow velocity and particle-fluid trajectories in the different parts of the arterial tree is also not yet possible. 

For radioembolization, we hypothesize that the fixed centroluminal catheter position of the antireflux catheter, associated with a more constant downstream branch distribution, may improve the predictive value of scout dose administrations. An accurate prediction of the therapeutic microsphere distribution, and thus estimation of the tumor to nontumor microsphere uptake ratio, would enable personalized dosimetry methods with the aim of achieving maximum tumor-absorbed doses without compromising treatment safety. Furthermore, the more homogeneous downstream branch distribution may improve tumor targeting. However, in-vivo validation of our findings is still required, because our surrogate model cannot replicate the complexity of clinical reality.

IO360: Will you be doing any further related studies?

van den Hoven: Yes. We have started a prospective clinical trial to investigate catheter-related effects in vivo.² This study has a within-patient randomized controlled trial design. In short, we will treat a total of 25 patients with unresectable and chemorefractory colorectal liver metastases with Ho166 microspheres, which were developed in our center to improve the imaging characteristics of radioactive microspheres used for radionuclide therapies. All patients undergo 2 procedures on the same day: a pretreatment procedure in the morning during which a scout dose of  Ho166 microspheres is administered followed by the treatment procedure with a therapeutic dose of  Ho166 microspheres in the afternoon. 

The use of the antireflux and standard catheters is randomly allocated to 1 of 2 selective injection positions (left and right hemi-liver). SPECT/CT imaging is used to assess the microsphere distribution after both procedures and determine the absorbed radiation dose in tumors and in the healthy liver tissue in the left and right hemi-liver. Eventually we will compare the predictive value of the Ho166 scout dose distribution, tumor to nontumor ratio, and post-treatment tumor response between the liver territories treated with the antireflux and standard catheters.

IO360: Anything else you’d like to share about the study and results?

van den Hoven: A manuscript of our study has recently been accepted for publication. I hope that it will trigger others to consider catheter design, particle-fluid dynamics, and related imaging technology as topics for future research or product development.

Editor’s note: Disclosure: Dr. van den Hoven has completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. He reports no disclosures related to the content herein. 

Suggested citation: Ford J. Andor van den Hoven, MD, describes a particle-fluid dynamics study for radioembolization of colorectal liver metastases. Intervent Oncol 360. 2015;3(7):E83-E88.

References

1. Jernigan SR, Osborne JA, Mirek CJ, Buckner G. Selective internal radiation therapy: quantifying distal penetration and distribution of resin and glass microspheres in a surrogate arterial model. J Vasc Interv Radiol. 2015;26(6):897-904. 

2. Surefire Infusion System vs. Standard Microcatheter Use During Holmium-166 Radioembolization (SIM). ClinicalTrials.gov Identifier: NCT02208804. https://clinicaltrials.gov/ct2/show/NCT02208804