Skip to main content

DARPin Targeted Magnetic Hyperthermic Therapy for Glioblastoma

Final Report Summary - DARTRIX (DARPin Targeted Magnetic Hyperthermic Therapy for Glioblastoma)

Executive Summary:
DARTRIX, ‘DARPin Targeted RX (therapy)’ is a multidisciplinary collaborative project aimed at developing targeted hyperthermia as a treatment for glioblastoma. Heat is directly toxic to cancer cells and can also help activate the immune system within the tumour microenvironment. The approach is innovative and there is great need for new treatments as glioblastoma is virtually incurable and most patients die within 12 months of diagnosis. The project employed Designed Ankyrin Repeat Proteins (DARPins) which are small, stable, non-immunoglobulin human protein scaffolds that bind specific targets with exceptionally high affinity. The goal was to couple DARPins to superparamagnetic iron oxide nanoparticles (SPIONs). SPIONs can generate heat when stimulated by an alternating magnetic current and, by functionalisation with DARPins, the SPIONs would have potential to bind selectively to tumour cells.

The consortium brought together a range of diverse skills to achieve the ambitious aims of the DARTRIX project. DARPins were generated for glioblastoma-related targets and the lead DARPin was manufactured to standards of GMP (good manufacturing practice) with a unique cysteine for site specific attachment to SPIONs. A new range of GMP compliant SPIONs were created with excellent heating properties and conjugation strategies were optimised to link the DARPins to these particles. Full toxicity testing was conducted on the lead SPIONs which were shown to be well tolerated, with no adverse effects either systemically or at the site of administration. Methods were developed to reduce unwanted uptake of SPIONs by the reticuloendothelial system. DARPins were engineered with long hydrophilic random coil polypeptides, creating second-generation DARTRIX particles with superior specificity for targeting cancer cells. Furthermore, employing immunocompetent pre-clinical models of glioblastoma, the consortium demonstrated that the GMP-compliant SPIONs generated effective and localised heat in tumours using a bespoke medical device to induce magnetic alternating current hyperthermia (MACH). A scaled up MACH system was developed for clinical use and tested in a hospital environment. This scaled-up MACH system is now operating at clinically relevant field strengths and is undergoing testing to achieve CE Marking. Finally, novel ways to track DARTRIX particles in vivo, and to measure the biological response to particle-mediated hyperthermia within the tumour microenvironment, were developed in pre-clinical models.

The results generated during the DARTRIX project will contribute substantially to knowledge in the scientific field in addition to providing a sound platform to facilitate a future DARTRIX first-in-human trial.
Project Context and Objectives:
1 Context (State of Art at the project start, Key challenges)

DARTRIX is a multidisciplinary collaboration to develop therapeutic, biocompatible, targeted SPIONs for localised hyperthermia treatment of cancer. In current clinical practice, SPIONs are used as contrast agents in magnetic resonance imaging (MRI). However, when subjected to an alternating magnetic field, SPIONs can readily heat to temperatures above 42°C. The aim of the DARTRIX project was to harness the heating potential of SPIONs to develop an innovative new cancer therapy that would safely deliver effective localised hyperthermia within the tumour. Challenges were to generate biocompatible GMP-compliant SPIONs that would be safe and effective in vivo and to build a clinical-grade MACH system capable of generating therapeutic heat in patients. Because SPIONs are not tumour specific, the ultimate goal was to target them to cell surface proteins such as epidermal growth factor receptor (EGFR), a widely recognised tumour target that is often overexpressed in glioblastoma. DARPins were chosen as targeting agents due their favourable affinity, stability, small size, economy of production and amenability to protein engineering. The DARTRIX project focused on treatments for glioblastoma (WHO grade IV), a highly infiltrative, rapidly progressive primary brain tumour that continues to be associated with a poor prognosis. There is no effective standard treatment for management of patients with recurrent/relapsed glioblastoma and an urgent need for new treatment approaches.

2 Objectives of the project

The DARTRIX project aims to develop a new, safe and efficient approach to targeted hyperthermia treatment by creating and applying a cancer-targeted SPION or ‘DARTRIX particle’, in parallel with a custom-made MACH system to deliver the alternating magnetic field. The DARTRIX particle should be clinically compatible, reproducible, targetable to appropriate tumour areas, imageable in patients, non-toxic prior to MACH activation and able to produce measurable localised heat upon activation. A successful outcome of the DARTRIX project would form a foundation for development of new treatments for glioblastoma and eventually other diseases.
Project Results:
1. UZH

The challenge of the project was to couple targeting moieties to magnetic nanoparticles, which would survive heating, be compatible with many kinds of coupling chemistry, show extreme specificity and be non-immunogenic, allow the attachment of additional labels and help the final particles to be resistant to uptake by the reticuloendothelial system (RES).

We chose to use Designed Ankyrin repeat Proteins (DARPins) as a basis. DARPins are small, non-immunoglobulin human protein scaffolds that have great potential to form the targeting moiety of clinically effective therapeutics. UZH has developed a strategy to build combinatorial libraries of repeat proteins by extracting sequence and structural information from compatible natural repeats to design a consensus amino acid sequence motif encoding self-compatible repeat modules. Indeed, DARPins are ideally suited for this particular therapeutic purpose — they are readily generated to bind specific targets with exceptionally high-affinity and they are remarkably stable, even at high temperatures. Furthermore, DARPins can be manufactured in gram quantities using simple bacterial expression systems amenable to GMP production and to economic scale-up.

Since the project needed DARPins to develop treatments for glioblastoma, the epidermal growth factor receptor (EGFR) was the most obvious target. EGFR is overexpressed in 50-60% of glioblastomas, the prototype anti-EGFR DARTRIX particle was functionalised using high affinity EGFR1-binding DARPins.

The anti-EGFR1 DARPin chosen was one which had no pharmacological activity (it did not compete with EGF on the receptor). It could be well expressed in E. coli, and its expression could be well transferred to P. pastoris in a GMP process. The DARPin carried a cysteine for versatile coupling to suitably modified magnetic nanoparticles.

Importantly, rigid quality control of the yeast-produced DARPins showed that they are fully functional and have the exact expected mass, and the previously established binding constant, as measured by surface plasmon resonance.

Important considerations in engineering the DARPins were to ensure target recondition when the DARPins were coupled to particles and to minimise uptake by RES and local macrophages. For this purpose, DARPins were engineered to be equipped with a long polar extension, either negatively charged, or uncharged (between 300 and 900 amino acids). A highly efficient expression and purification method was established for these constructs, and minor impurities were removed, using two purification tags, one of them cleavable and a very inexpensive affinity column (itself exploiting DARPins).

The so purified DARPins with cysteines and long polar extensions could be efficiently coupled to magnetic nanoparticles and were shown to specifically recognise the target on the surface of tumour cells.

In addition, DARPins that bind new targets relevant for glioblastoma were generated from the library by an extensive series of ribosome display experiments and subsequent characterization. E.g. a series of anti-cMET DARPins were generated and characterised in detail. They were shown to cover many different epitopes on cMET on different domains, show high specificity, react on cells and show low nanomolar KDs. Using the technologies established for the EGFR binders, which are completely generic, they can now be manufactured in very similar formats.


2.1 Development of DARTRIX particles

MICROMOD has introduced a new type of iron oxide dextran composite nanoparticles (perimag®) into the DARTRIX project. These new particles are produced in a clean-room facility for GMP compliance and show favourable and consistent heating properties under the conditions of the MACH system (RCL). The surface chemistry of these particles was tailored for site specific conjugation of DARPins with a terminal thiol group by introduction of carboxylic acid or amino groups to obtain DARTRIX particles. For surface functionalisation with amino groups, the dextran surface of the particles was cross-linked with epichlorohydrin followed by reaction with ammonia. Amino group density on the particle surface was altered by changing the general reaction parameters. Thus, DARTRIX particles with high and low density of amino groups were prepared to study the influence of the density of functional groups and surface charge on the cell-uptake of the final DARPin conjugate at UCL (Figure 1, see attachment). The surface potential of the functionalised particles was characterised by electrophoretic measurements of the zeta potential of the particles in dependence on the pH value. The DARTRIX particles with COOH groups on the surface possess a negative zeta potential over the whole pH range. The increase of the density of amino groups is accompanied by a shift of the zeta potential into positive direction (Figure 2, see attachment).

Different strategies were established to conjugate DARPins on aminated or carboxylated particles. The aminated particles were functionalised with maleimide groups for reaction with the thiol group of the DARPins. Hereby the length and flexibility of the spacer between particle surface and terminal maleimide groups was varied. The performance of rigid cyclohexyl spacers and flexible PEG spacers with 6, 12 and 24 ethylene glycol units was compared. Furthermore the density of DARPins on the particle surface was varied (Figure 3, see attachment). For DARPin conjugation on carboxylated particles, the COOH groups were activated with carbodiimide to facilitate reaction with a PEG spacer that containsterminal amino and maleimide groups (Figure 4, see attachment). All conjugation strategies were performed with the specific E69 DARPin and the He-G3 DARPin as reference.

UZH has introduced longer spacers between the bioactive E69 unit and the terminal SH group to achieve a higher flexibility of the DARPins on the particle surface for an improved specific cell binding of the DARTRIX particles. MICROMOD has obtained these new PAS- and XTEN-DARPin derivatives from UZH (Figure. 5, Table 1, see attachment) for conjugation to perimag® particles. The PAS-DARPins were prepared with a short and long flexible chain between the DARPin and the terminal cysteine.

The XTEN288-DARPins have a higher number of negative charges compared to the XTEN864-DARPins. Each PAS- and XTEN-DARPin was provided as specific E69 DARPin and as Off7 reference DARPin. These new DARPins were conjugated to maleimide functionalised perimag® particles according to Figure 6, see attachment. The size distribution and the zeta potential of all DARTRIX particles were measured by dynamic light scattering. The DARPin content of the DARTRIX particles was determined by a modified BCA (bicinchoninic acid) assay. The physico-chemical data of the obtained DARTRIX particles are summarised in Table 2, see attachment. All DARPin particle conjugates show a good colloidal stability and were provided for cell-uptake studies at UCL In addition the maleimide functionalised particles were reacted with cysteine as reference for the in vitro cell uptake studies at UCL.

All new DARTRIX particles had a hydrodynamic diameter of about 140 nm. The DARPin concentration of the particles lies in the range of 16-18 μg/mg Fe for the PAS-DARPin derivatives and of 17-21 μg/mg Fe for the XTEN-DARPin derivatives.

The conjugation of aminated perimag® particles with PAS-DARPins leads to a significant shift of the zeta potential from about +15 mV at pH=7 to negative values in the range of -10 to -15 mV (Figure 7, see attachment). The zeta potential of all XTEN-DARTRIX particles is more negative than that of the PASDARTRIX particles. This correlates well with the higher number of negative charges of XTENDARPins in comparison to PAS-DARPins (Table 1, Figure 8, see attachment). The isoelectric point of all new DARTRIX particles lies at pH=4.

All PAS- and XTEN-conjugated perimag® together with the cysteine-labelled reference particles were sent to UCL for in vitro cell uptake studies. Initial results from UCL demonstrated a high specific binding of PAS-E69 and XTEN-E69 particles in comparison to the corresponding Off7 conjugates and the cysteine reference.

DARTRIX particles with different iron concentrations, surface functionalisations and formulations were delivered in sufficient amounts for toxicity studies to UCL and TOPASS. The colloidal stability of these particles was investigated for particle formulations in water and artificial cerebrospinal fluid (aCSF). Dynamic light scattering measurement did not show any significant changes of the size distributions over a period of 6 months. The heating parameters of these particles were measured at RCL. The Intrinsic Loss Power (ILP) values of all particle types that were delivered for toxicity studies are in the range of 6.3 - 6.7 nHm²/kg.

2.2 Quality management system

MICROMOD has been registered as an ISO 9001 and ISO 13485 certified company. This registration includes all organisation processes applicable to purchase order processing, manufacturing, testing and quality assurance. These certificates attest MICROMOD a quality management system which allows us to be able to produce and develop medical products according to applicable legal or regulatory requirements respectively. These legal requirements are effective in addition to the technical requirements of the products.

The legal requirements for manufacture and distribution of medicinal products and active substances using as starting material are derived from the EU GMP guidelines. Part I of GMP guide describes GMP principles for the manufacture of medicinal products, Part II for active substances used in starting materials and Part III GMP related documents. The GMP guide is supplemented by a series of annexes. Compliance with the guide is regulated in various national laws (AMG, AMWHV, national and international pharmacopoeias). The requirements to investigational medical products are described in Annex 13 of the GMP guide.

MICROMOD was audited by UCL for verification of compliance with ISO 9001:2008 and ISO 13485:2015 and adequate good laboratory practices for manufacture of IVD. The quality system and operational activities for production, laboratory, QC & support operations in the context of manufacture of SPION for first DARTRIX clinical trial were in the scope of audit. QMS and facilities of MICROMOD were found to be compliant with current regulations and standards required for the manufacture of IVD.

3. RCL

3.1 MACH system

The MACH system (Magnetic AC Hyperthermia) was developed as a scalable and modular system that could be used for in vitro and in vivo as well as clinical applications. One of the key features of the system is the ability to self-tune to the resonance of the field applicator, in this case the head coil, which allows high efficiency of the transfer of energy from the system to the coil. Further to this, the self-tuning mechanism allows positioning of the coil to ensure maximum field exposure is applied to the region of interest i.e. site of the magnetic nanoparticles. Throughout the project the core PCB (printed circuit board) underwent 18 revisions, with each iteration making improvements on performance, scalability and safety. Figure 9 and Table 3, see attachment, highlight some of the major revisions to the PCBs along with maximum AC magnetic field achieved with the clinical coil.

3.2 In vitro / in vivo MACH development

Several lab scale systems were built for the DARTRIX project to undertake particle characterisation and small animal work, both of which used single PCB set-ups to generate the magnetic field. The in vitro / characterisation system utilised a multi-turn solenoid style coil that was capable of generating fields up to 15 kA/m at ~1MHz, whereas the in vivo / animal system used a split pancake Helmholtz style coil with a maximum field of 6 kA/m at 800 kHz. The open style in vivo coil allowed direct line of sight for infra-red thermography, something that other hyperthermia systems struggle with due to lack of access, particularly when using solenoid coils. Although the in vivo coil could only produce just under half the field of the in vitro model, considerable effort was made to ensure that the fields would be sufficient for animal experiments (see section on bioheat modelling) as well as providing as uniform field distribution as possible over a large treatment volume. In the case for the in vivo coil, the field homogeneity varied by only 10% over a 20x20x20mm volume, as shown in Figure 10 and 11, see attachment.

3.3 Clinical MACH development

The primary goal for device development was to build and demonstrate a magnetic hyperthermia system suitable for a clinical environment capable of generating an alternating magnetic field of 5 kA/m. The constraints of development lie mainly in portability and the ability to run the system using power from a 240V 13A socket i.e. not exceeding 3 kW of power. Early models showed that the scalability of the MACH PCBs would allow currents of 800-1000A to be generated in a single turn head shaped coil that would generate the required 5 kA/m. This required the use of 6x PCBs utilising the full 3 kW from a power socket. Although this style of coil would have sufficed there would have been little headroom to increase the field if required unless more PCBs were used along with a second power supply (thus limiting the portability of the system). To overcome this problem, a novel shaped coil was designed which was named the “dished elliptical coil”. This coil was a double turn ellipse that was dished over a cylinder to form a concave shaped coil that could then be positioned over the patient’s head, rather than round it. The added benefit here was the coil could now “project” the field more efficiently from the central part of the coil than a circular coil (Figure 12, see attachment).

This design also allowed for a second turn due to the decrease in total inductance per turn when compared with the single turn circular coil and thus reducing the required current to generate 5 kA/m from ~1000A to 600A. This structural change in coil design would permit some level of flexibility if higher fields would be required (although other considerations such as heat dissipation and high voltages would need to be addressed).

Another consideration to be made was in reducing the overall resistance of the coil. The impact of driving high frequency currents in a conductor is overcoming a problem called the skin effect. In short, the depth in which current can travel within a conductor is inversely proportional to the square root of the frequency. Therefore, the higher the frequency of the applied current, the less area the current can travel through the conductor i.e. less efficient. For a current at 300 kHz, the skin depth would be ~120 μm. It was therefore important to use a special type of wire called Litz wire that has hundreds of individually insulated strands per bundle.

The final version of the dished elliptical coil had an outer major axis of 19cm and an outer minor axis of 16cm dished over a 30cm diameter cylinder with a total number of 28 litz bundles each with 420 strands of wire i.e. 11,760 strands of individually insulated wires around the coil (Figure 13, see attachment).

3.4 Frequency Tracking

One possible advantage of the MACH system over a conventional hyperthermia system is the novel technique through which the system “tracks” the resonant behaviour of the coil. Firstly, this allows the system to precisely lock-in to resonant frequency and run at high efficiency without the need of user intervention to tune the drive circuit. Secondly, changes in the resonant frequency can be used to provide feedback to the user. It was noted that under controlled laboratory conditions it was possible to indirectly measure the temperature of magnetic nanoparticles through a change in resonant frequency. Further to this any sharp change in resonant frequency when positioning the coil can be linked to mass of magnetic material (Figure 14, see attachment).

3.5 EMC tests

With the MACH system able to operate at clinically relevant fields, the next steps will be to run a series of stability tests and fully document all aspects of the system which will include electromagnetic compatibility (EMC) testing with the final aim of obtaining certification that complies with the EN60601 standard for Medical Electrical Equipment and Systems. The goal of EMC is to make sure the equipment will function properly in an electromagnetic environment with emissions not exceeding existing standards as well as the equipment not being prone to failure due to radio frequency interference. Although there are over 70 standards that the MACH system will need to comply with in order to be CE-ready, we have been confident from early tests of the system beside medical equipment (ECG, Blood pressure and pulse monitors) with no signs of interference. Further to this the MACH system is purely an analogue system (in electronics terms) with no software or digital decisions required to control the system. From this perspective, it has been suggested that steps towards a CE-ready system will be easier to achieve.

Preliminary EMC measurements were performed at ETS (Electromagnetic Testing Services Ltd) on an operating MACH system driving the two-turn head coil. The system was running at 300 A peak, drawing about 500 W from the main supply (a higher power was not possible due to limitations in cooling facilities).

Emissions from systems are divided into two categories, conducted and radiated. Conducted emissions are measured in a lower frequency range, 150 kHz to 30 MHz, where the wavelengths are generally too long for radiation to be a problem. Radiated emissions are measured in a higher frequency range, 30 MHz to 1000 MHz (Figure 15, see attachment).

The system exceeded the limits for both radiated and conducted emissions. But given the lack of EMC precautions currently implemented in the prototype circuit, and the power levels being handled, the excess is acceptable. For instance, the highest radiated peak seen was around 57 dBμV/m at 3 m, which is equivalent to 0.15 μW if radiated isotropically, compared to the 500 W that the system is handling. It is also encouraging to note the relatively low level of emissions at the resonant fundamental (around 367 kHz), and at its low order harmonics, despite the huge currents we are generating at that frequency. These tests have allowed us to make a number of changes to the circuits to reduce the emissions and the modifications will be tested in-house with our own EMC test equipment to ensure regulations are met before applying for CE certification.

3.6 Bioheat thermal models

Aside from development of the hardware, RCL has put considerable effort into implementing bioheat models to estimate the required doses of magnetic nanoparticles for pre-clinical and clinical set-ups (Hergt & Dutz, (2007). The basic model, as shown in the equation (Equation 1, see attachment) has been modified to take into account the effectiveness of the heating of the magnetic nanoparticles (given by the ILP/SLP), the spread of material at the injection site and the heat conductivity of the tissue. Figure 16 shows a vast difference between achievable temperature rises for animal and clinical models using perimag® magnetic nanoparticles which is linked to the volume of injected magnetic material.

The limiting factor that prevents therapeutic temperatures in small animal models is the Fe toxicity limit (currently suggested as 0.6 mg for an average sized mouse). Whereas for the human model, the Fe limit is set at 150 mg and thus the required temperature rise can be estimated through the appropriate concentration and injectable volume. These Fe limits have been derived from the use of predicate materials such as Resovist and Feraheme, both of which are introduced into the body intravenously. The IFU for Resovist from Agis Commercial Agencies Ltd. 2002 state the following:

- The maximum dose tested in humans, 0.08 ml Resovist (equivalent to 2.24 mg Fe) per kg body weight, was well tolerated.

Which indicates that intravenous doses up to 2.24 mgFe/kg (or 40 μmolFe/kg ) have been tested in humans with no adverse effects. The recommended dose of Feraheme is an initial 510 mg dose followed by a second 510 mg dose 3 to 8 days later:

- Administer Feraheme as an intravenous infusion in 50-200 mL 0.9% Sodium Chloride Injection, USP or 5% Dextrose Injection, USP over at least 15 minutes.

Assuming 60 kg as an adult human’s minimum weight, the 510 mg dose corresponds to 8.5 mgFe/kg however this dose is considered as a “3 day dose” which in effect is equivalent to a daily dose of 2.8 mgFe/kg.

Other Fe-oxide predicate materials such as Sienna+ and NanoTherm are injected interstitially and in the case of NanoTherm offer much higher doses (35 mgFe/kg) based upon data in the literature (Johannsen et al., 2007). The rationale being that interstitial injection into the tumour would provide a higher level of safe Fe loading, yet the upper limit for circulatory Fe would be lower. Considering the worst case scenario whereby the injection is accidentally injected into the circulatory system, the upper limit for Fe to avoid toxic shock should be governed by intravenous limits and therefore for a 60kg human an Fe limit of 150mg was chosen. For the case of animal models, the Fe limit has been scaled according to a human equivalent dose based upon the conversion factors in Table 4 (see attachment).


The major challenge of the DARTRIX project for TOPASS was testing specifically designed iron oxide nanoparticles for hyperthermia in a preclinical setting that allows evaluation of their risk-benefit profile in terms of their potential application in patients with glioblastoma. Various test protocols were realised in mice and rats that enabled selection of specific candidate nanoparticles for further development. Specifically, we found that the surface charge of iron oxide nanoparticles is critical for their risk profile after intravenous application. In this regard particles with a negative surface charge are more likely to be tolerated compared to particles with a positive surface charge. This finding is consistent with observations using nanoparticles with other coatings but comparable sizes (Lutz et al., 2006; Thunemann et al., 2006; Cartier et al., 2007; Kaufner et al., 2007).

In addition, we demonstrated the fate of iron oxide nanoparticles after implantation into the brain. The bulk of the material deposited inside the brain will stay there for a long time. Using in vivo imaging technologies such as MRI or PET the localization of nanoparticles could be followed over time inside the body of living animals. These protocols helped to implement the 3R concept (Refine, Reduce, Replace) for the DARTRIX project. Figures 17 and 18, demonstrate MR images of DARTRIX particles in the brain and liver of mice (see attachment).

TOPASS will publish the results of the in vivo studies using DARTRIX particles. The DARTRIX project helped to move the nanomedicine community to new frontiers by expanding the border of our knowledge concerning the behaviour of nanoparticles in living animals, specifically iron oxide nanoparticles.

5. UCL

UCL’s major Scientific and Technical challenge was to develop a new bioprocess to manufacture DARPins that would be safe and effective as targeting agents for DARTRIX particles in the clinic. Initially we investigated manufacture in bacteria (E. coli) in accordance with published pre-clinical work on DARPins. However, after full evaluation we chose to use the yeast P. pastoris system as this allowed secretion of DARPin into culture broth, facilitating ease of obtaining pure product in a relatively simple process. Downstream processing of recombinant proteins produced by P. pastoris is easier than processing of recombinant proteins produced by E. coli, as only low concentrations of P. pastoris host cell proteins are released into the media. We worked closely with UZH in developing the process for E69 anti-EGFR DARPin and introduced a mutation to remove a potential breakdown site and improve yield of high integrity product. The process we developed was rapid and feasible to adapt to the rigors of manufacture to standards of Good Manufacturing Practice (GMP) in a quality controlled environment. This standard is a regulatory requirement for products used in patients.

5.1 Manufacture of GMP anti-EGFR DARPin

A 10L fermenter was used for process development and a 20L Clean in Place (CIP) machine for GMP. Figure 19 shows the complexity of fermentation apparatus required to provide a suitable controlled environment for the process. The high cell density of yeast culture is illustrated in the glass walled 10 Litre vessel on the left (Figure 19, see attachment).

An outline of the steps for upstream processing (USP) and downstream processing (DSP) is shown in Figure 20; each step was controlled by robust standard operating procedures (SOPs) written for the project. The final product was pure with no evidence of aggregation as illustrated in Figure 21, see attachment.

The final fill of E69 as an active pharmaceutical ingredient (API) was performed in a vertical laminar flow hood (VLFH) maintained in the GMP facility to a grade A standard (sterile environment). To ensure sterility and appropriate testing of the process the following SOPs were employed: use of biocides and alcohol to ensure a clean/sterile environment and use of special garments (see Figure 22, see attachment). Settle plates were in use to monitor sterility of the environment; Sabouraud Dextrose Agar (SAB) to detect fungal and Tryptone Soya Agar (TSA) for bacterial contaminations; upon incubation at 22.5 and 32.5 oC for 3 days, these plates will show fungal or bacterial growth in case there is a breach of sterility in the hood. In addition, finger dabs on TSA and SAB plates are taken from the operator. Finally, 1ml aliquots were prepared in cryo-vials, labelled, racked and placed in a monitored ultra-low temperature freezer (-80oC) (Figure 22, see attachment). The final product was tested according to set release specification and each test was reviewed by the QC and QA Managers. The product met all release criteria regarding: concentration, fill volume, product size & purity, aggregate analysis, endotoxin, host cell protein, sterility, EGFR binding and mass spectrometry and was qualified with a certificate of analysis (CoA).

5.2 DARTRIX Particle interactions with biological systems

In addition to overcoming challenges of GMP manufacture, UCL also developed in vitro and in vivo models to characterise interactions of DARTRIX particles with biological systems. This work included methods to reduce unwanted uptake of particles by the RES, as we initially published using near Infrared dyes (Abdollah et al., 2014) and later developed for radiolabelled imaging (Figure 23; manuscript in preparation - see attachment).

The large range of in vitro cellular models developed by UCL included glioblastoma cells, paired isogenic cell lines (positive or negative for EGFR expression) neural stem cells and macrophages. Confocal Scanning Laser Microscopy (CLSM) was optimised to visualise particle internalisation into these cells and together with electron microscopy/mass spectrometry the models provided powerful tools to analyse cellular/particle interactions (Figure 24, see attachment).

To evaluate specificity of DARTRIX particles, UCL employed fluorescence-activated cell sorting (FACS). The most recent application of this method was to demonstrate EGFR-specific uptake of new DARTRIX particles that were functionalised with DARPins bearing PAS or XTEN long polar extensions (Figure 25, see attachment).
Potential Impact:
1. UZH

1.1 Advantages displayed by the new materials by comparison with classical antibodies or Oligonucleotides

The strategy of the DARTRIX project contains a coupling step of the targeting agent to the magnetic nanoparticles. Furthermore, the nanoparticles need to be heated in the final treatment. These two requirements immediately show that highly stable, genetically modifiable binding proteins such as the DARPins have some key advantages. They can be genetically fused to very long unstructured regions, such that they have no steric hindrance to bind to cells. Furthermore, their high intrinsic stability make them resistant to heat-induced aggregation, a problem frequently encountered with many antibody fragments.

1.2 Clinical proof of concept and safety (lack of immunogenicity)

DARPins have been shown in clinical trials to not cause an immune response that would limit exposure, even after multiple injections over a prolonged period. This is due to the absence of aggregation (which would otherwise trigger a T-cell independent reaction) as well as the absence of T-cell epitopes in clinical constructs.

1.3 Scale-up and production methods

The scale-up of DARPins is very economic, and very robust against modifications of the molecules, such as additional cysteines or long unstructured extensions, as used in the present project.

2. RCL

The potential benefit to RCL lies with validation of a CE-ready medical device during first in man trials. It is also crucial that RCL can lay some fundamental ground rules for patient safety and tolerance to alternating magnetic fields. To date there is only one other company (MagForce, Germany) who have performed magnetic hyperthermia experiments, however the difference with RCL technology is the compactness of the device and the ability to apply the magnetic field with more focus to the region of interest. As far as RCL is aware there currently does not exist a CE Marked magnetic hyperthermia instrument for human use and based upon our progress through EMC tests as well as advice from our regulatory advisor we are confident that the equipment will be CE certified in the near future.

It is also worth noting that alongside development of the clinical system, RCL has gained insight into manufacturing characterisation and in vivo animal systems that have been used throughout the project. Pre-clinical data will play an important role in deciding on critical parameters for a clinical study such as field amplitude and exposure duration based upon magnetic nanoparticle dose. To this effect RCL has submitted a paper to the International Journal of Hyperthermia titled “Commentary on the clinical and preclinical dosage limits of interstitially administered magnetic fluids for therapeutic hyperthermia based on current practice and efficacy models”. Currently under review with a first round of minor corrections, the paper aims to review and suggest suitable dose limits for magnetic nanoparticles under a range of pre-clinical and clinical conditions. Once again, RCL is unaware of any peer reviewed guidelines for magnetic hyperthermia and aims to disseminate results gained from the project.


A very important initial result of the DARTRIX project for MICROMOD was the finding that perimag® particles provide high heating rates under the conditions of the MACH system from RCL. This makes the particles interesting for hyperthermia applications not only for the treatment of glioblastoma but also for other types of cancer in the future. The consistency of the heating properties of perimag® particles was clearly demonstrated. Within the DARTRIX project different surface modifications of perimag® were investigated regarding cell uptake and cytotoxicity. Perimag® with carboxylic acid groups on the surface were found to be non-toxic in in vivo mice studies up to a dose of 1.5 mg Fe per kg body weight. This information forms an essential framework to support our customers for the design of future medical applications.

The new generation of PAS- and XTEN DARPins from UZH was successfully conjugated to perimag® particles. A high specificity for target cells of the conjugates was demonstrated by in vitro cell uptake studies at UCL. The site-specific conjugation of perimag® to DARPins does not influence the biofunctionality of the DARPin molecules. This makes the perimag® particles a universal platform for the conjugation of different target molecules. These findings might strongly support the design of further investigations on glioblastoma therapy with hyperthermia.

MICROMOD has introduced perimag® as a standard catalog product into the market in January 2017. The advertisement for perimag® started at the 11th International Conference on the Scientific and Clinical Applications of Magnetic Carriers in Vancouver in May 2016 and was continued at the 7th International Workshop on Magnetic Particle Imaging IWMPI in Prague in March 2017.
The distribution will also be supported by the product presentation of perimag® at the 34th Annual Society for Thermal Medicine Meeting in Cancun in May 2017.

The mentoring of MICROMOD’s customers regarding customised conjugation of biomolecules on the surface of perimag® will be improved. The manufacturing of perimag® under clean room conditions will be offered at a high level of GMP compliance.

4. UCL

4.1 Clinical Impact

By developing a complete novel cGMP manufacturing system for DARPins using the yeast expression system, UCL’s work during the DARTRIX project will have clear impact on future scale-up and production methods for DARPins and other recombinant proteins destined for the clinic. Both the upstream processing (USP) and downstream processing (DSP) protocols are designed to be readily scaled-up and are streamlined in ways to improve cost of goods. For example, reduced effort in removal of contaminants, no requirement for centrifugation, cost effective chemically defined media and no use of animal derived products.

UCL will also contribute substantially to potential clinical impact of the project by exploiting findings in the DARTRIX project to develop treatments that are led by the needs of brain cancer patients. There is an urgent need for an enhanced program of clinical trials in brain tumour patients and theDARTRIX project will lead to a first in human clinical trial in glioblastoma patients. Currently, new therapies are first developed in other tumour types and only subsequently evaluated for use in brain cancer. Indeed, only 3% of brain tumour patients enter clinical trials (NCIN Routes to Diagnosis Report, published September 2012) and it is rare to have ‘first in human’ trials for brain cancer in the UK (Rampling et al. 2016). UCLs work, combined with other members of the DARTRIX consortium, will make strong contributions to the future treatment of brain cancer because information, knowledge and research outcomes from the DARTRIX project will underpin the clinical trials to take forward the new DARTRIX particles. These contributions include DARTRIX reported data on: favourable in vitro and in vivo heating potential in the MACH system, acceptable toxicity profile demonstrated in GLP studies, potential targetability of DARTRIX particles with a range of DARPins, stability and GMP compliance of DARTRIX particles. The strategic importance of this research into brain tumours is evident: in the UK alone approximately 9,000 people are diagnosed with primary brain tumour each year and there are almost 5,000 deaths. The majority of these deaths are attributed to a diagnosis of glioma. The current standard treatment following surgery in glioblastoma is 6 weeks of chemoradiotherapy followed by 6 months of temozolomide for those patients who have adequate performance status. For grade II and III glioma, there is marked improvement in overall survival with the addition of adjuvant systemic chemotherapy to radiotherapy (Buckner et al. 2016; van den Bent et al. 2013; Cairncross et al. 2013). However, the survival increase in glioblastoma is only modest with a median overall survival of 14.6 months (Stupp et al. 2005). There is no standard treatment for recurrent glioblastoma and options include best supportive care, nitrosureas or clinical trials. The median overall survival in recurrent glioblastoma is 5-10 months (Wick et al. 2010; Batchelor et al. 2013; Brown et al. 2016).

UCL will continue efforts to develop the DARTRIX concept for brain tumour patients before other patient groups. However, it is recognised that the DARTRIX project opens the gates to new treatments for other cancers and this will have a greater impact on public health. Indeed we have already opened negotiation to start UCL collaborations to take DARTRIX forward for patients with Melanoma and Head & Neck cancers. To date DARTRIX project work has secured funding from The British Council Russia Institutional Links (£145,951.00): Thermotherapy for the treatment of malignant brain tumours mediated by functionalised magnetic nanoparticles. The next stage will be to combine DARTRIX with immunotherapy investigating the role of therapeutic heating in reversing the hostile immune environment within a tumour. The rationale is strong: in the last decade, scientific and clinical research has demonstrated the ability of the immune system to control and potentially cure cancer and these ground breaking developments are driving a new era of therapeutic strategies for cancer treatment. We will build on our recent safety observations in the clinical use of immune checkpoint inhibitors in glioblastoma (Carter et al 2016). A further grant has been submitted to the “The Brain Tumour Charity: Quest For Cures” application call under the title of “Harnessing the Immune System in Glioma” in part, to take forward the work of the DARTRIX project. This grant application has been led by UCL.

4.2 Educational Impact

The DARTRIX project has been highlighted at two significant educational events with potential to influence future thinking; at “Antibodies: An evolving force in cancer treatment” (11 March 2015), the DARTRIX project was discussed in the context of glioblastoma treatment and future directions for research. The meeting was organised by UCL DARTRIX partners for the Oncology Section at the Royal Society of Medicine. It explored the current and future developments in the use of antibodies and DARPins for the diagnosis and treatment of cancer. There were approximately 70 attendees from diverse backgrounds, Medicine (Students to Consultants/Professors), Scientists (PhD students, Postdoctoral Fellows, Professors), Charity employees and Patent attorneys. A second Educational DARTRIX presentation was given as part of a Royal Society of Medicine Acute Oncology Education Day (15 January 2016).

The DARTRIX project funded a clinical fellow who has used their time to develop their doctoral research. In addition the fellow has been educated in the management of glioma patients and in clinical trial methodology. The fellow has developed a keen interest in glioma treatment and is ideally placed to take a leadership role in the future development of treatments for brain cancer patients.


Tumours arising in the central nervous system (CNS) make a significant impact on the health of the citizens of the European Union with approximately 27,700 new cases diagnosed each year (Crocetti et al. 2012). The DARTRIX project aimed to develop a new method of treatment for these cancers. By synergistic collaboration throughout the project, the consortium demonstrated complete success in bringing together resources that are not available in any single country. Thus addressing the recognised problem that cancer research is mainly undertaken at national level and tends to be considerably fragmented and diverse across the EU. During the DARTRIX project, DARPins were designed by UZH with a unique c-terminal cysteine for site specific attachment. UCL developed GMP-compliant methods for manufacturing the DARPins. MICROMOD generated a new range of DARTRIX particles, with surface chemistries tailored for attachment of DARPins. UCL and RCL developed methods for size fractionation of particles and determined the heating properties and UCL established methods to evaluate in vitro cell uptake and in vivo RES uptake of the DARTRIX particles. The consortium brought these skills together to achieve their ambitious goals. There is no single entity that has the academic, technical, clinical and commercial know-how to achieve this alone. Furthermore, private investors would be extremely unlikely support venture capital investment at this early stage although the potential benefits for participants are very high.

Abdollah MR, Kalber T, Tolner B, Southern P, Bear JC, Robson M, Pedley RB, Parkin IP, Pankhurst QA, Mulholland P, Chester K (2014) Prolonging the circulatory retention of SPIONs using dextran sulfate: in vivo tracking achieved by functionalisation with near-infrared dyes. Faraday Discuss 175 41-58.
Batchelor, T. T., P. Mulholland, B. Neyns, L. B. Nabors, M. Campone, A. Wick, W. Mason, T. Mikkelsen, S. Phuphanich, L. S. Ashby, J. Degroot, R. Gattamaneni, L. Cher, M. Rosenthal, F. Payer, J. M. Jurgensmeier, R. K. Jain, A. G. Sorensen, J. Xu, Q. Liu, and M. van den Bent. 2013. 'Phase III randomized trial comparing the efficacy of cediranib as monotherapy, and in combination with lomustine, versus lomustine alone in patients with recurrent glioblastoma', J Clin Oncol, 31: 3212-8.
Brown, N., C. McBain, S. Nash, K. Hopkins, P. Sanghera, F. Saran, M. Phillips, F. Dungey, L. Clifton-Hadley, K. Wanek, D. Krell, S. Jeffries, I. Khan, P. Smith, and P. Mulholland. 2016. 'Multi-Center Randomized Phase II Study Comparing Cediranib plus Gefitinib with Cediranib plus Placebo in Subjects with Recurrent/Progressive Glioblastoma', PLoS One, 11: e0156369.
Buckner, J. C., E. G. Shaw, S. L. Pugh, A. Chakravarti, M. R. Gilbert, G. R. Barger, S. Coons, P. Ricci, D. Bullard, P. D. Brown, K. Stelzer, D. Brachman, J. H. Suh, C. J. Schultz, J. P. Bahary, B. J. Fisher, H. Kim, A. D. Murtha, E. H. Bell, M. Won, M. P. Mehta, and W. J. Curran, Jr. 2016. 'Radiation plus Procarbazine, CCNU, and Vincristine in Low-Grade Glioma', N Engl J Med, 374: 1344-55.
Cairncross, G., M. Wang, E. Shaw, R. Jenkins, D. Brachman, J. Buckner, K. Fink, L. Souhami, N. Laperriere, W. Curran, and M. Mehta. 2013. 'Phase III trial of chemoradiotherapy for anaplastic oligodendroglioma: long-term results of RTOG 9402', J Clin Oncol, 31: 337-43.
Carter, T., H. Shaw, D. Cohn-Brown, K. Chester, and P. Mulholland. 2016. 'Ipilimumab and Bevacizumab in Glioblastoma', Clin Oncol (R Coll Radiol), 28: 622-6.
Cartier, R., et al. (2007). "Latex nanoparticles for multimodal imaging and detection in vivo." Nanotechnology 18(19).
Crocetti, E., A. Trama, C. Stiller, A. Caldarella, R. Soffietti, J. Jaal, D. C. Weber, U. Ricardi, J. Slowinski, A. Brandes, and Rarecare working group. 2012. 'Epidemiology of glial and non-glial brain tumours in Europe', Eur J Cancer, 48: 1532-42
Hergt, R., & Dutz, S. (2007). Magnetic particle hyperthermia—biophysical limitations of a visionary tumour therapy. Journal of Magnetism and Magnetic Materials, 311(1), 187–192.
Johannsen, M., Gneveckow, U., Thiesen, B., Taymoorian, K., Cho, C. H., Waldöfner, N., et al. (2007). Thermotherapy of Prostate Cancer Using Magnetic Nanoparticles: Feasibility, Imaging, and Three-Dimensional Temperature Distribution. European Urology, 52(6), 1653–1662.
Kaufner, L., et al. (2007). "Poly(ethylene oxide)-block-poly(glutamic acid) coated maghemite nanoparticles: in vitro characterization and in vivo behaviour." Nanotechnology 18(11).
Lutz, J. F., et al. (2006). "One-pot synthesis of PEGylated ultrasmall iron-oxide nanoparticles and their in vivo evaluation as magnetic resonance imaging contrast agents." Biomacromolecules 7(11): 3132-3138.
Rampling, R., S. Peoples, P. J. Mulholland, A. James, O. Al-Salihi, C. J. Twelves, C. McBain, S. Jefferies, A. Jackson, W. Stewart, J. Lindner, S. Kutscher, N. Hilf, L. McGuigan, J. Peters, K. Hill, O. Schoor, H. Singh-Jasuja, S. E. Halford, and J. W. Ritchie. 2016. 'A Cancer Research UK First Time in Human Phase I Trial of IMA950 (Novel Multipeptide Therapeutic Vaccine) in Patients with Newly Diagnosed Glioblastoma', Clin Cancer Res, 22: 4776-85.
Stupp, R., W. P. Mason, M. J. van den Bent, M. Weller, B. Fisher, M. J. Taphoorn, K. Belanger, A. A. Brandes, C. Marosi, U. Bogdahn, J. Curschmann, R. C. Janzer, S. K. Ludwin, T. Gorlia, A. Allgeier, D. Lacombe, J. G. Cairncross, E. Eisenhauer, R. O. Mirimanoff, Research European Organisation for, Tumor Treatment of Cancer Brain, Groups Radiotherapy, and Group National Cancer Institute of Canada Clinical Trials. 2005. 'Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma', N Engl J Med, 352: 987-96.
Thunemann, A. F., et al. (2006). "Maghemite nanoparticles protectively coated with poly(ethylene imine) and poly(ethylene oxide)-block-poly(glutamic acid)." Langmuir 22(5): 2351-2357
van den Bent, M. J., A. A. Brandes, M. J. Taphoorn, J. M. Kros, M. C. Kouwenhoven, J. Y. Delattre, H. J. Bernsen, M. Frenay, C. C. Tijssen, W. Grisold, L. Sipos, R. H. Enting, P. J. French, W. N. Dinjens, C. J. Vecht, A. Allgeier, D. Lacombe, T. Gorlia, and K. Hoang-Xuan. 2013. 'Adjuvant procarbazine, lomustine, and vincristine chemotherapy in newly diagnosed anaplastic oligodendroglioma: long-term follow-up of EORTC brain tumor group study 26951', J Clin Oncol, 31: 344-50.
Wick, W., V. K. Puduvalli, M. C. Chamberlain, M. J. van den Bent, A. F. Carpentier, L. M. Cher, W. Mason, M. Weller, S. Hong, L. Musib, A. M. Liepa, D. E. Thornton, and H. A. Fine. 2010. 'Phase III study of enzastaurin compared with lomustine in the treatment of recurrent intracranial glioblastoma', J Clin Oncol, 28: 1168-74.

6. Main Dissemination Activities

6.1. Publications (see attachment)

6.2.Conferences and presentations (see attachment)

6.3 Exploitation of results
As outlined in the initial description of work and reported throughout the project, the consortium made significant progress towards a future DARPin Targeted Therapy and has developed valuable IP. Through close monitoring and evaluating the progressing DARTRIX project results the consortium members were able to extend and strengthen their IPR portfolio for future exploitation. The handbook of Foreground IP has been updated accordingly to represent all IP generated throughout the project by all consortium members. While the dissemination of the “longer circulatory time for SPIONs through a clinical agent” and “fractionation of ferucarbotran” are being prepared through publications, UCLB is currently finishing the value assessment of potential patent filings (see attachment).

While the DARPin Targeted Therapy still requires further development, pre-clinical and clinical valuation, some of the generated IP is already being commercially exploited through individual consortium members. As outlined in previous sections, the DARTRIX collaboration has helped MICROMOD to generate crucial data for perimag® nanoparticles which are now not only a standard catalogue product but are also meant to be used in other customised medical applications outside of DARTRIX. While having extended the patent portfolio around the proprietary MACH technology, RCL has also made a significant step towards CE marking of the system. A current market review has also confirmed that there is currently no other system available which is able to deliver a locally targeted magnetic field like the RCL system. While completing EMC tests and other CE marking activities RCL will also start to commercially exploit and develop the MACH system.

While MICROMOD and RCL are already, or in due course, in a position to exploit some results of the DARTRIX project the consortium members are continuing to progress the technologies towards a personalised glioblastoma thermotherapy. As such the consortium members, led by UCL, are in the preparation of a potential first-in-man clinical trial which would not only deliver the first glioblastoma thermotherapy but also would further strengthen and validate the foreground IP generated during the DARTRIX project. Said follow-on research, which is necessarily to bring the technology into the clinic to benefit glioblastoma patients, will also incorporate and benefit from know-how generated i.e. use of specific size fractions, valuable pre-clinical data, potential use of clinical agents to increase SPION circulation time. That said, until the first thermotherapy in glioblastoma is delivered and the subsequent up take in health care systems is still several years away but the consortium members are continuing to work together to ensure that further down the line said exploitation is successfully managed.
List of Websites: