Skip to main content
European Commission logo
English English
CORDIS - EU research results
CORDIS
CORDIS Web 30th anniversary CORDIS Web 30th anniversary
Content archived on 2024-06-18

Image-guided pancreatic cancer therapy

Final Report Summary - IPACT (Image-guided pancreatic cancer therapy)

Executive Summary:
The demographic changes in Europe towards an aging society will make cancer the most important cause of death and morbidity. Currently 280 000 new cases of pancreatic cancer are diagnosed yearly worldwide representing less than 2.5% of new cancer diagnoses. However, as pancreatic cancer is age related, this malignancy is expected to become by 2030 the cancer with the highest incidence rate. Primarily diagnosed in late stage with limited therapeutic options, patients die with an average life expectancy of only a few months after diagnosis without any real therapeutic option. Within the IPaCT project a completely new therapeutic option is explored based on high intensity focused ultrasound, which allows non-invasive heating of deep-seated tissue to ablative and hyperthermic temperatures. To that aim, a novel US-MR-HIFU system that integrates HIFU with two imaging modalities, i.e. Magnetic Resonance Imaging (MRI) and diagnostic Ultrasound (US), ie. US-MR-HIFU, for image-guided thermal therapy of pancreatic cancer has been realized in this project. Furthermore, drugs encapsulated in temperature-sensitive liposomes (TSLs) haven been developed for localized, heat-triggered release within the pancreas. Efficacy of this approach has been shown in small animal studies. Finally, a Proof of Concept study has been performed where the US-MRI-HIFU system was used for thermal ablation of pancreatic tissue as well as for hyperthermia-triggered drug release of Dox and cisPt from TSLs. Both experiments were successfully concluded. The US-MR-HIFU system with its newly developed transducer was capable to deliver enough acoustic power at the depth of the pancreas to achieve multiple successful ablations. The double frequency matching also allowed to heat the pancreas to hyperthermic temperature to activate drug release from TSLs. For both drugs an enhancement factor was found in the range of 2-4 for cisPt and DOX respectively. The results obtained in IPaCT path the way to a new pancreatic cancer treatment based on HIFU. The results clearly justify further commercialization of the developed technology as well as TSLs for future clinical application.
IPaCT project brought expertise together form two technical SMEs on hardware (Imasonic) and software (Neagen), and an international MedTech company (Philips) represented by its subsidiaries in the Netherlands, Germany and Finland covering the entire technology development chain. UMCU covered knowhow on imaging and ultrasound technology and together with UHC medical application knowledge necessary for translation of the results into a potential clinical use. The partners TUE, LMU and UHC with their expertise on chemistry, material science and preclinical work successfully developed the TSL formulations including their scale up for studies in a porcine model and provided experimental evidence for efficacy of this approach. Finally, the expertise of UMCU, Philips Research and UHC on large animal studies allowed the joint PoC study performed at UMCU.
The following key results have been obtained:
• New transducer developed with longer focal length and improved acoustical characteristics and integrated US-imaging
• Integration of the transducer into a fully operational US-MR-HIFU prototype
• Development of several drug-TSL formulations (gemcetabine, irinotecan, cisPt) for local drg delivery including preclinical testing.
• Proof of Concept study showing successful ablation and image guided drug delivery of the pancreas in a large animal
• Results have been published in 11 peer reviewed journal articles with 5 articles currently under review and several articles still in planning
• Knowledge dissemination and outreach is reflected by more than 40 different activities reaching national and international audience

Project Context and Objectives:
The main objective of the IPaCT Project was to develop a new high intensity focused ultrasound system integrated in an MR scanner and equipped with diagnostic ultrasound imaging for image guided drug delivery for treatment of pancreatic cancer (US-MR-HIFU system). The latter also required the development of new TSL formulations with drugs applicable for pancreatic cancer thus leading to a technology centered research trajectory and a research arm focussed on development of TSLs and their preclinical testing.

Main objective 1: Development of a US-MR-HIFU system
• Development of a new transducer US-HIFU
• Development of electronics to interface with the US-HIFU system
• Development of software for image registration and therapy planning
• Integration of all components into a working prototype

Main objective 2: Development of suitable TSLs formulations
• Development of new TSLs using cisPt, irinotecan, gemcetabine as APIs
• In -vitro and in-vivo testing in small animals
• Upscaling of suited candidates for use in large animals

Main objective 3: Proof of Concept study
• Ablation of pancreas in a porcine model using US-MR-HIFU
• Hyperthermia induced drug delivery with TSLs in a porcine model using US-MR-HIFU

Objective 1 was addressed in WPs 1,2,3 covering transducer development, software development and system integration, while objective 2 was addressed in WP 4. Finally, WP 5 addressed several aspects that were necessary for translating experiments and technology development into large animal and potentially human applications. Those results, together with the prototype obtained from WP3 and the TSLs prepared in WP 4 were necessary to address objective 3, namely the final proof of concept study.

A new transducer design was pursued with 256 ultrasound elements based on Fermat’s spiral and Voronoi tessellation. As an additional requirement, the transducer had to be designed in a way to fit the existing mechanical positioning robot of a Sonalleve MR-HIFU system and had to include an ultrasound imaging transducer for additional ultrasound image guidance. Furthermore, as it had to operate in an MRI, all parts had to be MR compatible. In the first phase a mock up version was built to test production and some basic aspects regarding material choices and interconnects. In the second phase a fully operation transducer was built which was transferred for integration. Next, the entire software engine for image co-registration of MR and US imaging was developed. First, a new software tool for improved patient positioning based on imaging data was developed using the real-time pictures acquired by the ultrasound imaging transducer integrated into MR HIFU table. Optimal patient position is the one, that provides the best acoustic window from the transducer i.e. skin to the target tissue. Next, a fully automatic server process for co-registering the patient’s pre-operative and per-operative MR scans with other image data was developed. Furthermore, this new visualization server can provide the rendering of the wanted images in real-time in the common frame of reference. The integration comprised building the fully functional prototype of US-MR-HIFU. A main aspect of this activity was the design and realization of an electronic matching board that allowed operation at two frequencies of 750 kHz and 1.2 MHz. The lower frequency is preferred for hyperthermia while 1.2 MHz is preferred for ablation. The electronics had to ensure acoustic impedance matching at both frequencies for maximum efficiency. The electronic board was mounted into the existing table and interfaced with the US-HIFU transducer on the one side and with the electronic HIFU generators on the other side. Furthermore, software of the user interface had to be adapted to visualize the changed acoustic beam path. The US-MR-HIFU prototype was working according to specifications and expectations allowing ablations with more depth reach compared to the standard Sonalleve transducer while exhibiting an improved acoustic field.
In the project several new TSLs have been developed with active pharmaceutical ingredients being gemcitabine, cisPt, irinotecan and doxorubicin as a standard for comparison. The TSLs were based on DPPG2, which is a synthetic lipid as well as DSPC and DPPC. These TSL provide long circulation time and high stability at body temperature while providing rapid and quantitative release at hyperthermic temperatures. Several preclinical studies were performed that demonstrated efficacy in small animal tumor models for gemcitabine, irinotecan and dox-TSLs, experiments with cisPt are currently still ongoing. The best performing candidates, cisPt-TSL and dox-TSL were upscaled for the Proof of Concept study to show drug delivery in the pancreas of a large animal, i.e. porcine model.
Finally, the iPaCT project performed the Proof of Concept study using the US-MR-HIFU to demonstrate full operation in a porcine model. The latter was used as anatomical features and location resemble the human situation. First, ablation studies were performed which clearly demonstrated that the new HIFU transducer is capable to deliver enough energy at depth on more than 9 cm from skin surface, which was confirmed by pathology. Second, hyperthermia-induced drug delivery studies were performed using dox-TSLs and cisPt-TSLs. Even though hyperthermia could only be achieved for 6 and 8 minutes respectively, for both drugs an enhancement by a factor 2 and 4 was found demonstrating that this concept leads to improved drug concentrations in the pancreas. Sonication during these hyperthermia experiments were hampered by breathing induced motion. However, better sedation protocols should enable longer hyperthermia times which should translate in even higher achievable drug concentrations in the tissue.
Overall, the project met all objectives and delivers a US-MR-HIFU prototype ready to future commercialization as well as several drug-TSL candidates that could enter a path towards clinical use. Based on our results, we can confidentially state that the concept of image guided drug delivery using US-MR-HIFU is a viable option for treatment of pancreatic cancer. Due to the high costs and long development timelines of TSLs it is quite likely that first ablative treatments of pancreatic cancer will reach clinical trials before a drug delivery combination will be explored.




Project Results:
Overall Overview:

In order to address the overall objective of image guided drug delivery for treatment of pancreatic cancer, two lines of research were necessary within IPaCT. First, it was necessary to develop a high intensity focused ultrasound transducer with is suited to treat deep seated tissue with ablation and hyperthermia and integrate it with MR and diagnostic ultrasound for image guidance (US-MR-HIFU). Above activities were carried out in WP 1-3 comprising the development of the transducer (WP1), necessary software tools (WP2) and system integration into a working prototype (WP3). Second, hyperthermia (temperature) sensitive liposomes (TSLs) had to be developed for hyperthermia induced drug delivery with drugs applicable to pancreatic cancer. Scope of WP4 was the development of TSLs and their preclinical testing.
WP 5 gathered data that were necessary for translation to large animals of the technology and testing of some features of the prototype in human volunteers. Finally, a Proof of Concept study was performed in a porcine model, where all aspects came together, namely image guided drug delivery using new TSL formulation in a large animal using the US-MR-HIFU system.


Following is a description of the most important results per workpackage

WP1 – Technology Development of US-HIFU Transducer

The objective of WP1 involved the design and production of a new phased array high intensity ultrasound (HIFU) transducer on which an ultrasound transducer for diagnostic imaging was included (US-HIFU). In addition, design options for a Dual Mode TA US-HIFU matrix transducer (TA: Therapy + mode A imaging) was explored that allows application of therapeutic ultrasound and imaging using the same elements. The main deliverable of this WP is a fully functional US-HIFU transducer to WP 3 for integration into a complete system.

The work carried out in WP 1 has been divided over the following tasks:
• Task 1.1 : Specifications of linear arrays for the first US-HIFU testing (UMCU, IMAS, PHC, PEN)
• Task 1.2 : Design and manufacturing of the imaging linear arrays (Phase 1)
• Task 1.3 : Specifications of the integrated Dual Mode TAUS-HIFU transducer (Phase 2) (UMCU, IMAS, PHC, PEN)
• Task 1.4 : Design and manufacturing of the integrated Dual Mode TA US-HIFU transducer (Phase 2) (IMAS, UMCU)
• Task 1.5 : Exploration of Dual-Mode TB US-HIFU arrays
(IMAS)
WP 1 was headed by Dr. Sophie Gimonet, IMAS, with UMCU, PHC, PEN being participating partners.

Results and Innovation:
The purpose of WP1 was to integrate US imaging and HIFU functionalities into a single US-HIFU transducer as well as the exploration of dual mode array solutions. Prototypes have been designed and fabricated in two stages for integration in the existing MR-HIFU platform.
Phase 1: A commercially available ultrasound imaging array that fulfils all functional requirements has been identified. This probe was repackaged by IMASONIC to satisfy MR compatibility requirements and integrated in an existing Sonalleve V1. The imaging transducer has been clipped on the existing MR-HIFU matrix and first test of functionality have been performed in vitro on phantoms. Ultrasound imaging quality phantom that are typically used for ultrasound image processing tests were imaged and showed good resolution. Thus, this approach was pursued also for phase 2, where the same ultrasound imaging transducer had to be integrated on the newly designed 256 element channel phase array transducer.
Phase 2: an integrated Dual Mode TA US-HIFU transducer (HIFU matrix of 256 elements including a phased array imaging probe) based on an original concept, i.e. aperiodic paving named « VTFS » - Voronoi Tessellation and Fermat's Spiral was designed and fabrication processes were worked out. In parallel, solutions to improve the MR compatibility were tested, for example mock-ups were designed, built and tested to validate the principle of a revised design for the transducer, having in view the further expected needs for its integration in a clinical system. Using these insights, the 256 elements phased array was produced the ultrasound imaging probe was integrated with the HIFU transducer. The whole assembly was mounted in a housing has could be attached to the mechanical robot positioning system within the MR table. The characterization of the transducer using a hydrophone setup was according to prior performed computer simulations confirming the predicted high degree of focusing of the HIFU array in the sagittal plane and in the coronal plane. At 750kHz as well as at 1.25MHz the system (i.e. integrated Dual Mode TA US-HIFU transducer and dual-frequency matching board) is entirely diffraction limited and performed fully according to specification. Following PoC trials (realized in WP4), the ageing of the array has been studied. The impedance of the elements of the matrix have been analysed (their characteristic and their evolution). During next potential trials, it will be interesting to continue to monitor these data versus the excitation conditions used (for example: power level, tests duration...). Furthermore, simulations were performed in order to better understand the performance in imaging mode that could be expected from an array designed for HIFU generation such as the iPaCT TA US-HIFU transducer. These results were promising and confirmed the potential of the Dual Mode concept. In parallel, the dual mode capabilities of different structures - based on different technologies of realization - have been evaluated. Characterization results demonstrate that we are on track to find a solution to combine HIFU functionality with imaging capabilities. The refinement of this concept is in progress to improve even more these performances.

Conclusion:
During the IPaCT Project, a new US-HIFU Transducer was designed and produced that allows hyperthermia and ablation of deep seated tissue. The transducer operates at two different frequencies optimized for hyperthermia at 750 kHz and 1.2MHz or ablation. Integration of ultrasound imaging allowed motion tracking of moving organs. The final Proof of Concept study demonstrated full operation and showed high acoustical power output at focal length of 16 cm which was sufficient for pancreas ablation. The transducer can serve as a prototype ready for commercialization.


WP2 – Software for Image Analysis and Therapy Planning

The work package 2 focused on developing technology and solutions to support the two elementary phases of US-MR-HIFU therapy: preplanning and treatment monitoring. Software tools for integration of US imaging with MR Imaging had to be created to handle motion of organs and of the target tissue, while dealing with the complex acoustic environment of the human abdomen that is dynamically changing due to respiratory movement of the organs. An essential part of the therapy monitoring is to ensure that the acoustic path from the ultrasound transducer to the target is maintained for safe and efficient treatment. We will integrate the intraoperative ultrasound and MRI imaging with the pre-planning model to enable the operator to monitor the safety and efficiency during the procedure.

The WP 2 structure along below tasks:
Phase 1: Software tools for patient positioning and pre-operative planning
Task 2.1: Guide the positioning of the patient
Task 2.2: Improve the pre-operative planning
Phase2: Software tools for online monitoring of the MR HIFU therapy
Task 2.3: Methods for co-registering the image data from various modalities
Task 2.4: Ultrasound based motion tracking of a moving organ of interest
Task 2.5: Acoustic coupling monitoring and alerting in case of failure

WP 2 was headed by Lasse Jyrkinen (NEA) with contributions from PEN, UMCU and PHC.

Results and Innovation:
During the project, a functional software prototype was developed consisting of a new imaging console that provides enhanced complementary functionality to the standard Philips MR-HIFU planning and treatment console. The main new ideas were to use the integrated real-time ultrasound imaging and image co-registration of the multiple MR & CT studies for advanced visualization.
In Phase 1, a new software tool for improved patient positioning was developed. The innovation was to use the real-time ultrasound imaging sensor integrated to MR HIFU table for finding the optimal position of the patient. Optimal patient position is the one, that provides the best acoustic window from the transducer i.e. skin to the target tissue. Integrated real-time ultrasound is the excellent tool for finding the optimal position for the best acoustic window.
In Phase 2, a fully automatic server process for co-registering the patient’s pre-operative and per-operative MR and CT studies to the common visualization frame of reference was developed. Furthermore, the new visualization server can provide the rendering of the wanted images in real-time in the common frame of reference. The challenging problem of automatic co-registration of the abdominal MR to CT studies was successfully solved.
In Phase 2, an advanced algorithm for the ultrasound-based motion tracking was developed. The algorithm enables the user to select several tissues (or organs) that can be tracked simultaneously based on the real-time ultrasound imaging data.
The motion tracking was further developed for the monitoring tool of the acoustic coupling between the patient and the HIFU transducer. The algorithm is calibrated while the patient is in steady position. Based on the calibration the algorithm calculates the optimal threshold for the movement detection. When the algorithm recognizes patient movement, it alerts and the sonication can be suspended until the patient is steady again.
The real-time acoustic coupling monitoring tool is the ultimate benefit of the real-time ultrasound monitoring. Without integrated MR-HIFU ultrasound imaging and advanced iPaCT image processing and visualization tools, the patient movement cannot be reliably detected.

Conclusion:
All the goals of the WP2 were achieved. The developed software prototype proved our ideas and concepts to be functional in a real-world application. The developed technologies are proven and ready for integration to a commercial HIFU system.
WP3 – System Integration of US-MR-HIFU

The objective of WP3 involved the definition of functional requirements on a system level and the subsequent system integration of a dedicated US-MR-HIFU device suitable for ablation and hyperthermia therapy of pancreas cancer. To study the benefits of the combined US and MR imaging for HIFU therapy already at an early state of the project, the work has been split in two phases. In Phase 1 an US-imaging transducer has been attached to an existing MR-HIFU device to enable combined US and MR images. In Phase 2 a complete US-MR-HIFU system has been developed with a fully integrated dual mode transducer.

The work carried out in WP 3 has been divided over the following tasks:
• Task 3.1. Functional requirement specification for US-imaging functionality of (Phase 1 prototype) (UMCU, IMAS, PEN, PHC)
• Task 3.2 Functional requirement specification for US-MR-HIFU hardware (Phase 2 prototype) (UMCU, IMAS, PEN, PHC)
• Task 3.3: Functional requirement specification for Software (PEN, PHC, UMCU, NEA, IMAS)
• Task 3.4: Integration and testing (phase 1) (PEN, PHC, IMAS, UMCU)
• Task 3.5: HW component development for US-MR-HIFU hardware (PEN, PHC, IMAS, UMCU)
• Task 3.6: US-MR-HIFU integration and testing (UMCU, PEN, IMAS, PHC)
WP 3 was headed by Dr. Pedro Rodrigues, PEN, with UMCU, IMAS, PHC and NEA being participating partners.

Results and Innovation:
Phase 1: After deriving requirements relevant for HIFU beam path imaging from the different work package partners and building of a suitable US imager fixture, executed commissioning tests during Phase 1 demonstrated that simultaneous MR-US operation was possible with a 2D diagnostic US imager integrated in the HIFU treatment bed without mutual interference between the two modalities, supporting the choice of a new US imager for further integration in the Phase 2 combined HIFU transducer. Next to patient positioning and beam path assessment also motion tracking of pancreas has been covered by the concluded, requirements. Hence a software framework has been created providing extended visualization capabilities and a plug-in service where gating and motion tracking algorithms can be further exploited during Phase 2. The software-based US-MR image combination has been deployed and with volunteers and various motion mock-ups tested. Interfacing with different sub-systems (Sonalleve HIFU, MR and Verasonics US data acquisition) has been demonstrated, as well.
Successful dynamic 2D-US and MR visualization of liver/pancreas region in a healthy volunteer has been shown combining the hardware and software solutions developed in the different iPaCT work-packages. Visual assessment of the combined images showed that US can be accurately co-registered with dynamic MR, enabling the possible use of US-only systems to track organ motion leaving the MR system available for thermometry monitoring during HIFU delivery.

Phase 2: After integration of the newly developed dedicated 256 channel HIFU transducer, incorporating an additional real time 2D-US imager, in a commercial Philips Sonalleve patient treatment bed an adequate electrical interface matching board has been developed. The first version was a single treatment frequency matching board with 256 channels optimized for ablation with 1.3 MHz. As a single frequency matching board its size and interfaces were the same as the original matching board, though replacement could be done without additional mechanical modifications. After calibration of the acoustic output power with radial force balance method, this intermediate system has been used for further development of motion tracking and gating algorithm software for combined US-MR imaging took place.
With support of a 16 channel functional VTFS matrix mock-up of the newly developed HIFU transducer – resembling the various element sizes of the new transducer and consequently their various impedances - a first 16 channel dual frequency matching network was designed and developed, in order to investigate different passive dual frequency matching network topologies. After selecting three topologies with their three orthogonal modifications and two mounting options mezzanine boards for the individual matching networks and a 42 channel back plane mock-up with interfaces to the newly developed 256 channel HIFU transducer and to the rigid-flex patch PCB - the interface towards the generator cabinet - has been designed and developed. After evaluating the different electrical and mechanical properties, by measuring the reflection coefficients per network, one electrical and mechanical topology has been chosen.
Subsequently a full 256 channel dual frequency version enabling both hyperthermia with 750 kHz and 1.2 MHz as well as ablation with 1.2 MHz was successfully developed. Additionally, another rigid-flex patch PCB was designed and manufactured to enable the movement of the generator interface to the bottom side of the backplane. This allowed with small mechanical modifications of the supporting structure the mounting of the entire matching network in the basement of the patient treatment table.
After integrating this new 256 channel dual treatment frequency matching network in the Philips Sonalleve clinical HIFU platform, the system underwent rigorous tests to demonstrate its readiness for the iPaCT PoC study. Acoustic metrology reconfirmed that at both treatment frequencies 750 kHz as well as at 1.2 MHz, the acoustic energy delivery system performs fully according to specification.

Conclusion:
The generated output energy at both frequencies is sufficiently high for the envisioned treatments hyperthermia and ablation of deep-seated tissue, a matching network approach of backplane with mezzanines needed for the new electrical requirements can be - even with its much increased volume - in the field mounted. The originally envisioned replacement kit for a Philips Sonalleve clinical HIFU platform consisting of a HIFU transducer with its matching board is feasible.


WP4 – Technology Development of US-HIFU Transducer

The objective of WP4 was the development of new TSLs, including their in vitro and in vivo testing for pharmacokinetics and efficacy. Along that line, all necessary assays were to be established for drug quantification, toxicity tests and biocompatibility. The main objective within the WP was the Proof of Concept study, where ablation of a healthy pancreas had to be demonstrated in a porcine model as well as HIFU induced drug delivery both using the prototype US-MR-HIFU.

The work carried out in WP 1 has been divided over the following tasks:
• Task 4.1 : Development of new TSL formulations (LMU, UHC, PEN)
• Task 4.2: Pharmacological testing of temperature induced drug delivery in small animals
• Task 4.3 Feasibility studies of US-MR-HIFU for HIFU therapy monitoring in animals. (UMCU,PHC, PRH, PEN)
• Task 4.4 Large animal study of HIFU mediated drug delivery to the pancreas
(UMCU, PH, PRH, PEN, UHC, LMU
WP 14 was headed by Dr. Holger Grüll (TUE, UHC) with LMU, UMCU, PEN being participating partners.

Results and Innovation:
In the first phase of the project, the partners LMU and TUE/UHC worked on the development of new TSL formulations with active pharmaceutical ingredients being irinotecan, gemcitabine, cisPt, and DOX. In addition, other drugs such as SN38 were investigated but found unsuited as API for encapsulation within TSLs. DPPG2-based liposomes showed long circulation times and provided high plasma levels in in-vivo studies. Efficacy test were performed using rat models for gemcitabine, irinotecan and DOX-TSLs using hyperthermia induced local drug delivery. As hyperthermia devices local heating with light sources were used at LMU and MR-HIFU at TUE/UHC. All studies showed rapid release of the drug pharmakon in-vivo leading to high local concentrations with enhancement factors between 4 – 20 depending on the used protocol.
Milestone studies were the DOX-TSL trial performed by LMU in feline sarcoma patients showing efficacy in a spontaneous occurring tumor. Another key study was performed by PEN, TUE and UHC showing local drug delivery with DOX-TSLs in a rat tumor model. The novelty of this study was the combination protocol of MR-HIFU hyperthermia with local TSL-based drug delivery followed by ablation. With that approach the drug enhancement was highest also outperforming all other treatment protocols by efficacy. The use of paramagnetic liposomes, i.e. TSLs filled with a drug and with an Gd-based MR contrast agent, allowed to use MR to visualize and quantify drug release during MR-HIFu hyperthermia. The change of T1 contrast is proportional to the release amount of drugs. For real time image guidance, parallel imaging sequences were developed for the MR allowing interleaved acquisition of T1 and temperature maps.
Several studies were performed to investigate any adverse reactions to the DPPG2 based lipid such as hemocompatibility assays, and also biodistribution assays for paramagnetic liposomes. The hemocompatibility assays showed that neither DPPG2 nor nDPPG2 showed a significant complement activation compared to the negative control (NaCl). Furthermore, there was also no significant difference between DPPG2 and nDPPG2 in complement activation via the SC5b-9 signaling pathway which is a strong indicator that these TSLs can be used in patients with good tolerance. For paramagnetic TSLs it was observed that the encapsulated MR contrast agent is first accumulated in clearing organs such as liver and spleen, but subsequently excreted over time. Therefore, toxicity risks due to reminiscence of gadolinium in tissue is low.
Finally, the formulations of cisPt and DOX in DPPG2-based TSLs were tested in a porcine model for blood kinetic and biodistribution. Both TSLs were administered by constant infusion over a time span of 30 minutes reaching peak plasma concentration of about 60-80 % ID/g. After stop of infusion, plasma concentrations decay exponentially due to clearance of TSLs and additional leakage of the API from its liposomal carrier. The leakage is higher for DOX compared to cisPt most likely due to a relaxation of the pH gradient that is used for loading of DOX.
Finally, the PoC study was performed consisting of an ablation study of healthy pancreas in a porcine model and, second, of two hyperthermia drug delivery studies using DOX-TSL and cisPt-TSLs together with the US-MR-HIFU system.
For the HIFU sonications, the pig is placed on a spacer developed by UMCU to compress bowel and stomach tissue and to create an acoustic window. Next, the sonication were planned using US-MR-HIFU where the US imaging is used to characterize the acoustic beam path. Ablation was performed at 1.2 MHz with 250 W acoustic power until the target temperature of about 60 ºC was reached. A CE MR scan showed the presence of an ablated, non-perfused tissue volume within the pancreas. After the US-MR-HIFU experiment, the pancreas was excised for later histological studies confirming the presence of a thermal lesion at the target area.
For hyperthermia induced drug delivery, the TSL solutions were infused of 30 minutes and while the peak reached its maximum, a hyperthermia treatment was performed heating a treatment cell of 1.4 cm diameter and roughly 4 cm length within the pancreas to 42ºC. Unfortunately, the animal returned to spontaneous breathing as sedation was not deep enough, leading to strong motion within the target region, which caused the automatic safety algorithm to stop further sonication. Thus, both hyperthermia treatments were only relatively short with 6 and 8 minutes. It should be emphasized that there is little doubt that technically hyperthermia could be performed for a longer time span, in case a better anesthesia regimen is used. However, even for these short hyperthermia times, an enhancement factor of DOX of 4 was found while the enhancement factor for cisPt is around 1.5-2.

Conclusion:
From the work presented in this WP, we conclude that DPPG2-based TSLs can be translated into clinical use as they are well tolerated and show promising pharmacokinetic properties. The DPPG2-based TSLs are a carrier system that allows to encapsulate a variety of different APIs that were investigated in this project. Efficacy of hyperthermia induced local drug delivery was shown in several small animal experiments. In large animals, TSLs were well tolerated providing reasonably high plasma levels around 30 minutes after start of perfusion. Most importantly, the final Proof of Concept study demonstrated that the US-MR-HIFU system with its new transducer is capable to ablated deep seated tissue at depth of 9 cm from skin layer and beyond. The system is also technically ready to heat target tissue for extended time periods to hyperthermic temperatures. Our studies show that even short hyperthermia times can already achieve significant high drug levels in the heated organs.
Our PoC study demonstrated that HIFU ablation and hyperthermia induced drug delivery is a viable treatment option for pancreatic cancer as soon as TSLs are clinical available.


WP5 Technical Feasibility of US-MR-HIFU for Clinical Translation

The main objective of WP 5 was to translate technical research on image guidance of HIFU therapies into a possible clinical setting by performing evaluations on human volunteers. Specifically, new MR thermometry methods and protocols were developped and evaluated in volunteers, imaging protocols of the pancreas were developed and tested as well as motion. The main objective was to test and use ultrasound imaging of the new US-MR-HIFU system on volunteers for evaluation of beam path obstruction and for assessment of target motion.

The work was carried out along 4 tasks:
• Task 5.1 – MR thermometry of the pancreas
• Task 5.2 – MR perfusion imaging and (tumour) tissue characterization of the pancreas
• Task 5.3 – Location and motion of pancreatic tumours
• Task 5.4 – US navigation for motion detection and motion compensation of MR imaging of the pancreas
• Task 5.5 – US A-mode imaging for detection of beam path obstructions
WP 5 coordinated by Dr. Clemens Bos (UMCU), with PEN, PRH, PHC as contributing partners.

Results and innovation:
1) MR thermometry of the pancreas
Studies on the feasibility of PRFS-based thermometry of the pancreas within iPACT has resulted in advances in the area of subject preparation and field drift correction.
It was shown that MR thermometry in the pancreas can be considerably improved if the duodenum is filled with fluid, such that susceptibility artifacts from air tissue transitions are substantially reduced. In addition, several designs of dome shaped model acoustic coupling devices were evaluated for minimizing the air-tissue interfaces in the beam path, as well as shortening the distance the soundwaves have to travel in tissue. It was found that a large (approx. 20 cm disk of 2 cm high was appropriate for making the pancreas head and body acoustically accessible a prone human subject. Moreover, it was shown that such a shape did not interfere with MR thermometry. For respiratory compensation, it proved most effective and robust so far to limit data acquisition to the exhalation phase.
For applying mild hyperthermia for prolonged durations, correction of B0-field drift is required. We demonstrated that interleaved acquisitions, here could serve to inject FID acquisitions, from which the B0-drift could be estimated. Initially, we had expected to be using field-probes for this application, however, without the need of additional hardware, the phased-array coil elements were shown to provide useful FID signals that could serve for 0-th order drift correction, but provided sufficient coil elements were available, also for spatially resolved drift correction.
Selected methodology was transferred to Task 4.3.1 for the preclinical Proof-of-Concept study in large animals.
2) MR perfusion imaging and (tumour) tissue characterization of the pancreas
It was demonstrated that ASL of abdominal organs was feasible, also in the pancreas. In addition, the feasibility of interleaved acquisition of relaxometry maps and MR thermometry using PRFS was investigated. Inserting a PRFS acquisition hardly influenced the relaxometry results, and could even contribute to its precision by providing a time window during which magnetization can effectively recover. Moreover, relaxometry often includes its own normalization/reference data and is largely based in signal magnitude, thus allowing per treatment tissue characterization. Insertion of relaxometry, however, significantly affected PRFS thermometry. Careful investigation allowed us to demonstrate that this effect was due to a disruption of the eddy current steady state. Subsequently, it was shown that by characeterization/calibration of long-term eddy currents and then taking this information into account in the eddy current compensation algorithm, the negative of these steady state disruptions on PRFS could be mitigated.
Pain secondary to pancreas cancer is thought to be transmitted via the celiac plexus, a structure of nerves behind the pancreas. Hence, the celiac plexus, in palliative treatments is a frequent target for ablation. Within the iPACT project we designed an MR neurography technique, based on flow sensitized signal dephasing, that allows to depict the structures of the celiac plexus in high detail, thus e.g. permitting targeting for ablative therapies.
3) Location and motion of pancreatic tumours
The motion of the pancreas in prone positions was evaluated in healthy volunteers using 4D MRI information. Motion was extracted from the 4D MRI using an optical flow based method. This information was then exploited for planning a novel treatment strategy, where the energy delivery continues for the complete respiratory cycle. This improves the duty cycle of the HIFU treatment, and could substantially reduce the treatment times (in terms of number of sonications required) depending on the size of the target lesion.
US navigation for motion detection and motion compensation of MR imaging of the pancreas
Images from the integrated US imaging transducer in the iPACT system were used to track the liver tip using optical-flow based displacement estimation. The pancreas was located slightly too deep for the stand-off imaging design that was used in the iPACT system. Hence, using the built-in imaging transducer, a motion estimation signal could be generated that could be fed back into the system for synchronizing the MR acquisition to the motion. An interface to this effect was designed, but we faced issues during implementation that led us to pursue other priorities within the project.
4) US A-mode imaging for detection of beam path obstructions
An obstruction detection method was designed, based on generating a cavitation cloud using high pressure, and then detecting the cloud using A-mode imaging. However, this method deposits significant energy in the subject and was therefore not further evaluated in volunteers.

Conclusion:
WP5 has delivered advances in MR thermometry in the abdomen, especially focused at temperature monitoring over prolonged periods of time, such as hyperthermia. Motion of the pancreas was characterized in treatment position, and a planning approach implying that motion was designed. Finally, novel methods for functional and anatomical evaluation of the target area at the pancreas were developed and evaluated and the potential of MR-US tracking was demonstrated.


Potential Impact:
In 2013 the iPACT consortium set out on a mission to advance the knowledge of applying combined MR and US-guided HIFU in the pancreas, specifically by triggering intratumoral release of anti-cancer drugs from temperature sensitive liposomes. In 2018, a prototype MR-US-HIFU system was available, that contained innovative components, invented and designed during the IPaCT project. With this prototype successful animal studies were performed demonstrating the capability of the system to locally heat tissue in the pancreas, and to either ablate tissue or trigger local drug release of cisplatinum or doxorubicin so as to increase the local concentration in the treatment area.
Besides the technological successes and learnings, which will be covered in more detail below, the intensified interaction of engineers with clinicians involved in pancreas cancer care, has led to a sharper identification of challenges and opportunities. Late detection is one of the key issues for the poor outcomes with pancreatic cancer, but at the same time treatment options should become available for the patients that at time of diagnosis do not fulfill the requirements for surgery, currently the only potentially curative treatment. HIFU remains a candidate option, especially now that with the iPACT project, in addition to work by other institutions in this domain, a number of approaches to gain non-invasive access to the pancreas have been proposed and have clinically or preclinically have been evaluated. As such, the IPACT members are being recognized internationally for their contributions to technology and concepts for treatment of pancreas cancer. In early 2019, the FUS foundation has invited Holger Grüll (UHC) and Chrit Moonen (UMCU) as panel members for an expert discussion on bringing forward HIFU treatment of pancreatic cancer in Bethesda.
iPaCT impact on the Industry
Within the IPaCT project, a Dual Mode US-HIFU transducer (HIFU matrix of 256 elements including a phased array imaging probe) was implemented, based on an original concept, i.e. aperiodic paving named « VTFS » - Voronoi Tessellation and Fermat's Spiral. This paving contributes to minimize grating lobes and so reduces unwanted effects outside the targeted area. Hence, it helps to preserve tissues outside the treatment zone, for example close to the transducer, where the ultrasound energy enters the body.
Besides for delivering HIFU energy to the pancreas, this transducer concept could be applicable to abdominal applications, e.g. liver, kidney, as well as peripheral applications, e.g. thyroid, breast, or beyond such as in the brain, for example. The transducer was designed to work in both modes, viz. HIFU ablation and hyperthermia. The bandwidth is sufficiently broad to work at two frequencies (750 kHz-1.2MHz) with large power levels.
To Imasonic, a high-tech SME from Besançon the knowledge and know-how gained in IPaCT is highly valuable and research and technological development are presently continued in order to exploit this knowhow:
• by means of adapting the technology for the development of transducers for various applications, notably for abdominal therapy (including with Profound Medical and for equipping the Sonalleve system), brain therapy and other therapeutic applications with European and non-European system manufacturers.
• in further research and demonstration of Dual and Multimode functionalities combining imaging and therapy
• in further adaptations of the technology making it compatible with other treatment modalities like histotripsy.

In order to use the transducer in both modes, a dual band matching board was designed that needed to be MR compatible and fit in a confined space. Thus, a matching with a minimum number of inductors and minimum coupling between the matching of individual circuits was built. The dual band matching board addresses the need for using a broadband transducer system that may be applied for several therapeutic approaches. Currently, HIFU systems tend to be very application specific, which leads to hardware and investment duplications, limiting the clinical application spectrum of an installation. The iPACT dual band system has thus provided approaches for increasing the versatility of HIFU systems, which could be of high interest for vendors of HIFU equipment such as Insightec, Profound, IGT, Theraclion.
In addition, software to guide the iPaCT MR-US-HIFU system was implemented as a demonstrator, which was able to visualize, register and plan a HIFU treatment. The activities, have provided Neagen Oy, an SME from Finland with know-how and multimodality registration, as according to their CEO Lasse Jyrkinen:
“Alhough the HIFU application itself is outside of the short-term commercial interest of the Neagen Oy, we have learnt a lot on relevant image processing algorithms. Particularly, we have been able to make an extensive research on the implementation of the modern image co-registration methods and algorithms.

We foresee to develop image co-registration to several clinical applications to take our neaLink enterprise imaging system to the next level. We have already developed support for image co-registration to our server platform. The first clinical results are naturally co-registering of the MR-MR, CT-CT and CT-MR image studies. We are working towards intelligent workflow improvements for the diagnostic radiology. Image co-registration plays there a key role.”
Many of the MR imaging guidance innovations in IPaCT used the capability of Interleaving MR methods with fast switching between different acquisition types or imaging contrasts. For example alternating between relaxometry and MR thermometry, or spectroscopy acquisitions and MR thermometry. IPaCT helped to showcase the potential of this approach for interventional guidance, but also application in general purpose radiology can be imagined, since the flexible combination of different contrasts will be important to increase speed, robustness and accuracy in MR imaging applications related to screening, diagnosis, therapy and follow-up. Therefore, the concept of interleaving MR scans will potentially be integrated into future Philips MRI product SW releases.

An important step towards the required technological readiness level for product integration was reached during IPaCT by closely analyzing possible interactions of quickly interleaved sequence types and applicability of combining two MR parameters (T1/PRF temperature and T2/PRF temperature) during a therapeutic intervention. The observed high impact of eddy currents from interleaved sequences on the signal phase is e.g. relevant for exact temperature mapping applications as pursued in IPaCT.

Interleaved scanning is expected to play a role in speeding up overall examination time in MR, contributing to robustness by sensing physiology and motion, as well as for supporting advanced therapy guidance strategies. Currently, Philips is MR imaging supplier for promising 3rd party image guided therapy systems such as the MR-LINAC (Elekta Unity) and MR-HIFU (Profound Sonalleve and TulsaPRO).
Also in the domain of thermosentistive liposomes, the work at University Hospital Cologne and at the Klinikum of the University of Munich within the IPaCT project has had substantial impact. It was shown that thermosensitive liposomes (TSL) based on DPPG2 with encapsulated chemotherapeutic drugs as developed during the IPaCT project are quite promising for the treatment of patients with pancreatic cancer in combination with HIFU, aimed at achieving hyperthermia. They will lead to higher intratumoral drug concentrations and thereby potentially improve antitumor activity without increasing systemic toxicity. The technology platform of DPPG2-based TSL (see Figure 3) is currently under development by Thermosome GmbH, an SME from Munich, for clinical application. A future perspective for the TSL with encapsulated chemotherapeutic drugs will be the further evaluation regarding scale-up and storage stability for potential clinical use with an adapted formulation.
Impact on the academic sector
The University hospital of Cologne (UHC) is one of the largest academic hospitals of Germany and covers all medical specialties. Currently, UHC has about than 12000 employees and still growing with investments into new buildings and equipment. UHC is one of the few German comprehensive cancer centers with pancreatic cancer being one of its focus areas. The department of radiology at UHC decided in 2015 to invest into MR-HIFU for treatment of uterine fibroids as well as bone metastasis and furthermore, to extend the application ranage of HIFU towards new oncological applications. As a consequence, Prof. Dr. Holger Grüll received an appointment as full time professor, before at the Eindhoven University of Technology, to lead the MR-HIFU group. The work performed at UHC within the framework of IPaCT created much awareness of that technology at UHC and also at the pancreatic cancer center of Cologne leading to joint activities. One outcome of IPaCT is the laboratory for translational MR-HIFU research which was established at UHC and allows large animal studies. Furthermore, a startup company of UHC, Soluxx GmbH, engaged into prototyping of MR-HIFU equipment and adjacent tools, for example gel pads, positioning devices that will be needed for future human applications. UHC is now applying for funding to perform a clinical phase 1 study for ablation and hyperthermia in pancreatic cancer, translating the results into clinical use.
Thanks to IPaCT, UHC is now the leading HIFU center of Germany, which is reflected by steadily increasing patient numbers for MR-HIFU applications such as ablation of uterine fibroids. Future dissemination of the results comprise application for a MR-HIFU Center at the German Research Society. On educational level, UHC is now training visiting medical doctors from interested hospitals to get familiar with MR-HIFU from theoretical point of view as well as practice. For the latter, hands on training course have been established. A dedicated course on MR-HIFU will be offered within the curriculum of medicine as well as for natural science students starting form 2019/20 onwards.
At the University Medical Center Utrecht (UMCU), IPaCT has strengthened the research in the technical and image guided therapy domains, with innovations such as the VTFS transducer design, image registration applications, MR sequences for tracking and methods for perfusion imaging. The IPaCT project contributed to the mass of the HIFU activities at UMCU, leading to follow up funding in the domains of HIFU neurology applications, and MR perfusion imaging, and intensified collaboration with radiotherapy, gasteroenterolic surgery and anatomy at UMCU.
Moreover, the IPaCT project, through strengthening the relations between the participating beneficiaries as well as the scientific advisory board members was pivotal for setting up the FURTHER consortium that successfully applied for H2020 funding in 2018 (FURTHER: Focused Ultrasound and RadioTHERapy for Non-invasive Palliative Pain Treatment in Patients with Bone Metastasis), with UMCU and UHC. Clinical efficacy results and healthcare economic evaluations from FURTHER will be pivotal in securing reimbursement for HIFU treatments. They will pave the way for other, more complex future applications of HIFU in Europe, by securing a business model with which to sustain clinical HIFU infrastructure, and thus allow advancement of the HIFU field in its entirety.
Additional H2020 and national applications between the partners are in preparation.
Finally, the IPaCT project has successfully trained early stage researchers (ESRs) for the European technology job market. 2 PhD’s have been awarded and 2 more PhD’s are near completion. These ESR are now employed by a company commercializing temperature sensitive liposomes, developing hybrid imaging/therapy systems in an academic setting, or will continue as a PostDoc on the afore mentioned H2020 FURTHER project.

List of Websites:
At the beginning of the project a public website has been established which can be found at http://www.IPaCT.eu . The public part gives an overview of the objectives of the project and is used to inform the public about the project. The consortium only part is used for internal communication, storage of submitted deliverables and reports, exchange of data and reporting within the project.

The consortium consisted of 6 partners from 4 countries, where the partner Philips participated with three different legal entities, namely Philips Oy, Philips Research Netherlands (PEN), Philips Research Hamburg (PRH). Due to legal changes, the latter is later mentioned as Philips GmbH. Furthermore, two SMEs, Imasonic (IMAS, France) and Neagen (NEA, Finland) participated and the four academic institutions, Eindhoven University of Technology (TU/e, The Netherlands), the University Medical Center Utrecht (UMCU, The Netherlands), the Ludwig Maximilian University (LMU, Germany) and the University Hospital Cologne (UHC, Germany).
Below list is far from comprehensive and lists key personnel from each partner participating in IPaCT.

Holger Grüll (UHC, TUE; Overall and PI UHC; holger.gruell@uk-koeln.de)
Patrick Voihs (UHC,Project coordinator; Patrick.voihs@uk-koeln.de)
Juan Castillo Gomez (UHC; participant; Juan.castillo-gomez@uk-koeln.de)
Lukas Sebeke (UHC; participant; Lukas.sebeke@uk-koeln.de)
Clemens Bos (UMCU; main contact, WP leader; c.bos@umcu.nl)
Chrit Moonen (UMCU; Scientific coordinator; C.Moonen@umcutrecht.nl)
Mario Ries (UMCU; participant; m.ries@umcutrecht.nl)
Elsbeth den Boer (UMCU; Project Financial Signatory; e.j.denboer-4@umcutrecht.nl)
Lars Lindner (LMU, main contact, scientific coordinator; Lars.Lindner@med.uni-muenchen.de)
Barbara Wedmann (LMU, Participant, Barbara.Wedmann@med.uni-muenchen.de)
Susanne Lindemann (LMU; Project Financial Signatory; susanne.lindemann@med.uni-muenchen.de)
Sophie Gimonet (IMAS; WP leader, main contact; sophie.gimonet@imasonic.com)
Gerard Fleury (IMAS, Main contact; Gerard.fleury@imasonic.com)
Jean-Luc Guey (IMAS; Participant; jeanluc.guey@imasonic.com)
Siozic Courtel (IMAS, Project Financial Signatory; soizic.courtel@imasonic.com
Lasse Jyrkinen (NEA; Main contact, scientific coordinator, Project Financial Signatory;lasse.jyrkinen@neagen.com)
Edwin Heijman (PEN, Philips Scientific coordinator; Edwin.heijman@philips.com
Ernst Hermens (PEN; Participant; ernst.hermens@philips.com)
Patrick Keur (PEN; Project Financial Signatory; patrick.keur@philips.com)
Rainer Zitzman (PEN; participant; rainer.zitzmann@philips.com)
Jochen Keupp (PRH, Philips; Main contact; scientific coorinator; jochen.keupp@philips.com)
Steffen Weiss (PRH, Philips; participant; Steffen.weiss@philips.com
Leon Luwijs (TUE; Project Financial Signatory; l.f.luwijs@tue.nl)
Jan van Hout (TUE; Project Financial Signatory; j.p.v.hout@tue.nl)
final1-ppp3-summary-figure.pdf