Skip to main content

Natural sensory feedback for phantom limb pain modulation and therapy

Final Report Summary - EPIONE (Natural sensory feedback for phantom limb pain modulation and therapy)

Executive Summary:
Background. Amputation of a limb may result from trauma or surgical intervention. The amputation traumatically alters the body image, but often leaves sensations that refer to the missing body part. In 50-80% amputees, neuropathic pain develops, also called phantom limb pain (PLP). Both peripheral and central nervous system factors have been implicated as determinants of PLP. Also, PLP may be triggered by physical (changes in the weather) and psychological factors (emotional stress). Recent evidence suggests that PLP may be intricately related to neuroplastic changes in the cortex, and that these changes may be modulated by providing sensory input to the stump or amputation zone. However, the understanding of why PLP occurs is still poor, the basic research results have not been tested on a large scale in the clinic, and there are no fully effective, long-term treatments readily available on the market. We therefore aimed to challenge the status-quo of PLP therapy by offering technological solutions that will invasively or non-invasively induce natural, meaningful sensations to the amputee to restore the neuroplastic changes in the cortex and thereby control and alleviate PLP.
Objectives. O1: Through coordinated multi-centre clinical trials, we aimed to assess the effectiveness of providing invasive/non-invasive sensory feedback (direct peripheral nerve stimulation/mechanical pressure or electrical stimulation) with or without simultaneous operation of hand prosthesis device for controlling phantom limb pain. O2: We aimed to assess the associated cortical neuroplastic, psychological and cognitive components of pain. O3: We aimed to provide clinical guidelines. O4: We aimed to build novel innovative technological pre-industrial systems for delivering invasive/non-invasive sensory feedback based on existing solutions emerging from previous funded EU research.
Technical developments. Route 1: Direct peripheral nerve sensory feedback. The invasive technology was planned to be tested at three hospitals in Europe. Multiple transverse, intrafascicular electrodes (TIME-4H) were implanted in the median and or ulnar nerves of volunteer amputee subjects. Electrical stimulation was delivered through the active sites by the multi-channel, miniaturized electrical stimulator placed outside the body. The TIME-4H electrodes were surgically removed after completion of the study. Route 2: Non-invasive sensory feedback. The non-invasive technology was tested at five universities in Europe and the USA. We applied mechanical sensory feedback (i.e. air pressure) through silicone pads, or electrical stimulation through of-the-shelf electrodes with or without including a hand prosthesis device. Across technologies: To deliver the electrical stimulation sequences and also to obtain quantitative and qualitative measures on the effect of the microstimulation, a semi-automatic and computerized platform was developed (Psychophysical Testing Platform).
Clinical trials outcomes. Route 1: Direct peripheral nerve sensory feedback. Two upper limb amputee volunteers participated in to receive ‘direct’ peripheral sensory feedback. Route 2: Non-invasive sensory feedback. 29 volunteer subjects enrolled in the non-invasive trials, including upper limb amputees, lower limb amputees and subjects with complex regional pain syndrome (CRPS). Across technologies. The group analysis indicated that there was, on average, a significant reduction in the phantom limb pain. Upper extremity subjects showed the largest effect as a group (as measured by the VAS and NAP of VAS measures). The lower limb amputees did not show a change in pain (as measured by the VAS and NAP of VAS measures), however they showed significant reductions in neuropathic pain (burning, paresthesia as measured by the NPSI) and reduction in pain interference on daily activities (as measured by the BPI). Contraindication is suggested for subjects with Complex Regional Pain Syndrome (CRPS) as the therapy appeared to increase the paroxysmal pain in some CRPS subjects (too few subjects to reach a definitive conclusion). Due to the heterogeneity of the subjects, it was not possible to achieve conclusive, joint results on the cortical organization/re-organization.
Exploitation and impact. The consortium has built novel technological systems based on existing technologies emerging from previous EU funded research and demonstrated the technologies in clinical trials with overall positive results. Business model cases were developed and the IPR situation analysed as the framework for brining the technologies to market. In case of the invasive clinical trials it proved challenging to obtain approval by the relevant competent authorities throughout different European countries, and it was highly difficult to identify volunteer subjects for this part of the study. However, we believe that the proposed work has laid a foundation for translating the basic research results into real clinical market applications and thereby provide long-term, patient-specific solutions to a large group of patients suffering from phantom limb pain.

Project Context and Objectives:
Amputation of a limb may result from trauma or surgical intervention. The amputation traumatically alters the body image, but often leaves sensations that refer to the missing body part. In 50-80% amputees, neuropathic pain develops, also called phantom limb pain (PLP). Both peripheral and central nervous system factors have been implicated as determinants of PLP. Also, PLP may be triggered by physical (changes in the weather) and psychological factors (emotional stress). Recent evidence suggests that PLP may be intricately related to neuroplastic changes in the cortex, and that these changes may be modulated by providing sensory input to the stump or amputation zone. However, the understanding of why PLP occurs is still poor, the basic research results have not been tested on a large scale in the clinic, and there are no fully effective, long-term treatments readily available on the market.

Hypothesis (Figure 1). We aim to challenge the status-quo of PLP therapy by offering technological solutions that will invasively or non-invasively induce natural, meaningful sensations to the amputee to restore the neuroplastic changes in the cortex and thereby control and alleviate PLP. We will assess the effect of cortical neuroplastic, psychological and cognitive components of pain and integrate the knowledge into clinical guidelines.
Figure 1. EPIONE hypothesis

Grand objectives

Objective I: Treatment of PLP. No effective, long-term treatments are currently available for PLP. In addition, there is no consistent knowledge on which type of sensations may be effective in affecting the cortical plasticity, and the strategy for applying sensory feedback. We aim to directly compare two routes for providing a more long-term or permanent solution for the amputees. We aim to provide clinical guidelines that include a recommended protocol for delivering sensory feedback therapy, methods for assessment of the PLP, factors determining the degree of PLP experienced by the individual subject and the sensory augmentation method. Through coordinated, multi-center clinical trials, we will deliver and assess a non-invasive sensory feedback solution and an invasive sensory feedback solution that in the future will offer the ultimate, permanent, invisible and cosmetically acceptable interface. Both solutions may be integrated with the operation of a hand prosthesis which many subjects already use on a daily basis.

Objective II: Understanding PLP. The mechanisms underlying the painful perception of a missing body part are still unclear. We aim to investigate if generation of PLP can be explained by changes in the cortical map that follows amputation, and whether PLP can be quenched by restoring the cortical map. We aim to assess the associated cortical neuroplastic, psychological and cognitive components of pain.

Objective III: Innovative solutions. There are no dedicated medical technologies available on the market with the aim to provide sensory feedback to control and alleviate pain. We aim to build novel innovative technological pre-industrial systems for delivering invasive/non-invasive sensory feedback based on existing solutions emerging from previous funded EU research.
The methodologies and outcomes are summarized in Figure 2.

Figure 2. Methodological approaches to be investigated, prototype systems developed and assessment methods.

Organization of the work
The ultimate outcomes of the EPIONE project are to deliver clinical guidelines and prototype technologies for the use of natural sensory feedback for phantom limb pain modulation and therapy. Although the solutions offered by the consortium will be based on existing technologies and clinical protocols we aim to investigate multiple factors as possible determinants of PLP (e.g. cognitive and psychological) and assess multiple aspects of providing sensory feedback (e.g. different sensory modalities and technologies for delivering sensory feedback) that are unprecedented in the literature. As a consequence of this, it is believed that it will be important to evaluate the obtained results and possibly refine the clinical protocol and technologies during the project lifetime. To address the objectives and reach the defined milestones, the work is structured in three main, sequential phases, where a level of iteration is incorporated in the second phase. ‘

• Phase I: Design and development (month 0-12): In phase I, the preparatory work for the clinical trials will be carried out. A common clinical protocol will be defined for all clinical trials (across type of sensory feedback to be provided and type of technology to be used) (WP1). Technologies to deliver invasive/non-invasive sensory feedback will be adjusted / refined to meet the clinical protocol requirements and market needs (WP2+ WP3). To efficiently drive the EPIONE technologies towards the market, a preliminary market analysis will carried out during the first phase to map the market needs and requirements (WP6).
• Phase II: Test, evaluate and revise (month 12-38): A large portion of the work will be invested in phase II. After the necessary approvals for the clinical trials have been obtained, the first group of clinical trials will be carried out (WP4 + WP5) and results will be collected and evaluated across the different clinical trials (WP1). In parallel, a full market analysis will be carried out (WP6). Based on the preliminary scientific outcomes (‘lessons learned’) and the results of the market analysis, the clinical protocol and technologies may be refined before the second group of clinical trials is initiated. During the last part of this phase, specific exploitation strategies will be developed.
• Phase III: Deployment (month 38-48): In the final phase, all results will be integrated to deliver the clinical guidelines (WP1) and demonstrate prototype specifications and technologies (WP6)

Project Results:

WP1. Clinical Trials Governance

Partners involved: All

The overarching objective of WP1 is to oversee, harmonize and coordinate multi-centre clinical trials of multiple neuromodulatory treatments and evaluations of phantom limb pain (PLP) such that statistical evidence of their differential efficacy can be tested, and based upon this evidence develop and establish good practices guidelines of effective treatment and evaluation. The work package was broken down into four tasks. The first task aimed to establish, maintain, shepherd and refine a common clinical protocol to be used across the multi-centre clinical trial. The second task aimed to collect and evaluate the group results to measure whether the therapeutic arms of the trial resulted in a change in the phantom limb pain experienced by the subject. The third task aimed to collect the experiences from the clinical partners to retrospectively establish a common set of guidelines and practices for delivering the effective therapies for treating PLP. And, the fourth task aimed to observe, monitor and facilitate the theoretical ethical framework within which the clinical trial and its participants operated.

Overview of main results
• A common clinical protocol which guided the development of the testing platforms, and the common execution of the multi-center clinical trial was established. The common protocol, furthermore, defined measures for group analyses which were used to determine the therapeutic outcomes and evaluate their effect sizes.
• The group analysis from the trial indicates that there was, on average, a significant reduction in the phantom limb pain experienced by the amputee subjects that participated in the trial.
• Upper extremity subjects showed the largest effect as a group as measured by the VAS and NAP of VAS measures.
• Although lower extremity subjects did not show a change in pain measured by the VAS and NAP of VAS measures, they showed significant reductions in neuropathic (burning, paresthesia) pain as measured by the NPSI measure and reduction in pain interference on daily activities as measured by the BPI measure. This contradictory outcome comes with the added observation of an increase in prosthetic limb use. This leads us to hypothesize that the therapy reduced PLP to enable increase use of their prosthetic limb which resulted in an increase in non-neuropathic residual limb pain. Thus we believe that the increase in non-neuropathic pain masked the reduction in neuropathic pain.
• The therapy appears to have a strong positive effect in reducing phantom limb pain for subjects with amputated toes.
• Contraindication is suggested for subjects with Complex Regional Pain Syndrome (CRPS) as the therapy appeared to increase the paroxysmal pain in some CRPS subjects. Although there were too few CRPS subjects to reach a definitive conclusion, future experiments with CRPS subjects should proceed with caution.
• The therapy showed to have a short carry over, and thus analysis of the group was made in the period between baseline (i.e. before initiating therapy) and the final week of therapy (therapy week 4). Effect measured at the outcome showed some reversion in some cases, which reduced the average outcome of the therapy.
• All therapy arms as measured in the final week of therapy vs baseline showed, on average, a positive effect in reducing phantom limb pain.
• The ethical framework underlying the project and its execution was developed by an independent ethical advisor.

Details of Main results
Establish and adjust/refine clinical protocol
The protocol defined the inclusion/exclusion criteria, timeline and phases of the trial, minimum durations and composition of the sessions, and the self-report instruments that measured the magnitude and extent pain, change of phantom sensations, emotion, interference by pain on activities of daily living, and objective instruments that aimed to measure changes in the cortical representations of sensation. The protocol served as the roadmap for the site-to-site protocols that were submitted by each clinical partner for ethical approval of their respective clinical trial, as well as the specifications blueprint for the computer based psychophysical testing platform (a tool to systematically collect data) developed as part of the EPIONE project. Finally, the protocol defined the common instruments to be used to assess the change in PLP experienced conditioned by the therapeutic intervention arms across the multi-center trial.
The refinement of the protocol, which came between the defined Round 1 and Round 2 of clinical trials, extended the inclusion criteria to include other classes of subjects that experienced phantom limb pain (lower extremity amputees) and neuropathic pain (CRPS) to address recruitment difficulties of upper extremity unilateral amputees, loosened the calendaring rigidity to minimize the day to day intrusion on the subject’s lives, and increased the measurement frequency for measures that were susceptible to drop outs. The general timeline and integrity of the measurements were otherwise unchanged to enable compatibility of the two trials to a combined analysis. The successful application of the common clinical protocol ensured compatibility of the protocols conducted at the various clinical trial sites, ensured compatibility of the data collected at each site, and enabled a combined group analysis to be undertaken.

Collect, integrate and evaluate all results from the clinical trials
During the course of the project, six clinical trial sites ran the study following the common clinical protocol defined in Task 1. A total of 31 total participants were enrolled, 23 started the therapy phase and 20 completed the protocol. At the time of this report one participant is still undergoing therapy. The data of this final subject is not included in the group analysis, and heretofore not included in the participant counts.
The subjects included in the analyses were those who started the therapy phase. The subject pool consisted of 11 upper extremity amputees (UE), 8 lower extremity amputees (LE), 3 subjects with complex regional pain syndrome (CRPS), and 1 subject with all 10 toes amputated (BLT). Summary of the subjects, site, therapy modality and measures received by IU for group analysis are summarized in Table 1.
Preliminary analyses between groups indicated that upper extremity and lower extremity subjects responded differently based on the psychometric measures, thus the subjects were grouped into four diagnoses: upper extremity amputees, lower extremity amputees, CRPS, and bilateral toes amputee.
The 24-hour Visual Analogue Scale of Pain (VAS) and the Non-overlap of All Pairs (NAP) of VAS, as well as the total measures and sub-scales of the Neuropathic Pain Symptom Inventory (NPSI), Brief Pain Inventory (BPI), and Profile of Mood States (POMS) were analysed to calculate the difference in measures as a function of protocol phase (1-4 weeks in therapy, outcome and follow-up weeks) to quantify whether the therapy resulted in a measurable change in the nature of pain (VAS, NAP of VAS, NPSI), interference of pain in the subject’s activities of daily living (BPI), and emotional status (POMS).
Although fMRI was included as an objective measure, due to the grouping of the subjects into upper and lower extremity groups, a sufficient number of subjects, n<10, was not achieved to enable group analysis of the fMRI.
The overall results are summarized in Table 2, which shows the average change by group in the scores from the primary psychophysical instruments. These results were cross-validated using a longitudinal statistical measure, the Linear Mixed Effects and Tukey post hoc. The effect, and significance of the effect are indicated in Table 3.
Although the table only shows the difference in score change from baseline to the last week of therapy, the changes indicated here are indicative of the general overall change seen longitudinally.
The statistical significance of the effect size was estimated using the Linear Mixed Effects (LME) longitudinal analysis coupled to a within-design analysis of variance (ANOVA). In all cases, p<0.05 was used as the standard level of significance. A p-value ~0.1 was used as the threshold to indicate trending changes that could become significant with the inclusion of more subjects. The overall results from the psychophysical instruments are summarized in Table 3.
The table further shows the p-values from a between-design 1-way ANOVA and within-measures repeated measures (RM) ANOVA. The 1-way ANOVA does not link change to subject, and thus was not used. RM ANOVA was limited in possibilities to run post-hoc analyses, but was retained as a cross validation of the LME+ANOVA method used.

Table 1: Summary of subjects and group analysis measures for group analysis. The subject information includes the de-identified subject code, diagnosis, round site and mode of therapy. LE = lower extremity amputee, UE = upper extremity amputee, CRPS = complex regional pain syndrome, BLT = bilateral toes amputee. 31 subjects were recruited for the study. 7 subjects withdrew (W) before the therapy (Tx) phase. 1 subject’s data was lost, leaving 23 subjects who underwent the trial. 3 of these subjects withdrew during the therapy phase and 20 subjects completed the trial. The Notes indicate details on drop-outs or withdrawals of subject. W TxN for the 3 subjects who entered the therapy phase indicate the last week in which they received therapy (i.e. Tx2 indicates the 2nd week of therapy). The blocked out fields indicate the data set received for group analysis. Empty fields indicate missing data.
Table 2: Summary of the average change in primary psychophysical outcome measures instruments. This table shows the change in score at the last week of therapy from the baseline measure for all subjects that entered the therapy phase. The lower table is a legend showing the values for clinical effect thresholds. Green indicates positive change, while red indicates negative change

Table 3: Results of the statistical tests from the primary psychophysical instruments to measure pain (24hr VAS, 24hr NAP of VAS, NPSI), pain interference on activities of daily living (BPI), and emotional state (POMS), grouped by diagnoses. Results indicate the p-values of the change. A p-value < 0.05 was considered significant, and is highlighted in green if the direction of the change was a positive effect. Changes trending in the positive direction that did not meet the significance criteria are highlighted in yellow.

Task 3. Provide therapy guidelines
Based on the results and collected clinical observations from the consortium, a guidelines document was authored that provides a retrospective guidance on the delivery of the therapy and suggestions for best practices. The overall outcome from the trial indicated that all arms of therapy showed some degree of positive effect on PLP for the amputee subject groups (UE, LE, BLT). CRPS showed mixed but negatively trending results suggesting contraindication for this diagnosis. Given the positive outcome for amputees, the guidelines suggest that non-invasive methods be applied first to determine efficacy for the patient. In cases where there is limited residual limb for electrode placement and access to hot spots, the implanted approach could be indicated. Caution is indicated in assessing pain in lower extremity amputees due to the possibility of interference with residual limb pain resulting from increased use of weight bearing prostheses.

Task 4. Ethical issues management
The ethical considerations during EPIONE have been thorough and consistent throughout the whole project. The IEA has initiated, facilitated and executed several different activities.

The following regulatory activities have been carried out during the project EPIONE.
1. Acknowledgment of relevant laws, principals and regulations
2. Identifying the relevant ethical field, in moral philosophy as well as in medical ethics / bioethics
3. Checking up on all ethical approvals
4. Analysis of the applied procedures of informed consent
5. Observation of cooperation, and work flow
6. Interview with participating patient

The following advisory activities have been carried out during the project EPIONE.
1. Developing an adequate method for relevant ethical reflection
2. Identifying adequate moral concerns, theoretically thickened and practically recognizable
3. Discussing appropriate learning strategies
4. Questionnaire disturbed amongst the partners, and analysis of the survey
5. A lecture on ethics and medical ethics for the partners
6. Workshop session with the partners
7. Evaluation with EPIONE management

It is the opinion of the IEA, that the partners in the project EPIONE have clearly demonstrated a more than adequate ethical awareness, of principals, in character and in judgment.
Ethical guidelines, principals and regulations have been adequately identified and satisfactory applied. Issues of relational ethics have been addressed and acknowledge, and to some extent also soundly developed. The partners have demonstrated a respectful behaviour and sufficient moral concern in their interaction with the participating patients. The partners have been taught medical ethics and normative theory, and they have verified appropriate interest, knowledge and competencies in these fields. The partners have also competently taken active part in practised ethical deliberation relevant for EPIONE.

WP2. Technologies for non-invasive sensory feedback

WP leader: EPFL

The objective of WP2 was the development of several non-invasive sensory feedback strategies for humans, designed to alleviate the symptoms of phantom limb pain. Different types of non-invasive feedback systems were developed, including (1) non-invasive mechanical stimulation, (2) non-invasive electrical nerve stimulation (TENS) without prosthesis, and (3) TENS used in conjunction with a prosthesis. Furthermore, relevant steps needed to be taken to facilitate future commercialisation of the new systems.

Overview of main results
• A system for non-invasive mechanical stimulation of the skin was successfully developed, including a hand prosthesis with integrated sensors, a custom socket/liner with mechanical actuators contained within.
• A custom system for non-invasive TENS, integrating a commercial TENS stimulator (INOMED), a computer and a custom developed stimulation software.
• A system for TENS sensory feedback with a hand prosthesis, incorporating a commercial TENS stimulator (REHASTIM), a single board computer running a custom control software, a commercial robotic hand (PRENSILIA) and an sEMG acquisition device.
• Preparatory work on commercialisation of devices was carried out to facilitate further development and exploitation.

Details of main results
The main goal of WP2 was to prepare three research setups for further clinical tests, these included (1) a non-invasive mechanical stimulation device with a hand prosthesis, (2) a non-invasive TENS stimulation device with a hand prosthesis, and (3) an open-loop TENS setup. All three systems were successfully developed. Additional details regarding each system are introduced here.
First, a system capable of conveying sensory feedback via mechanical skin stimulation was developed. This system was composed of a prosthetic hand with incorporated sensors. The readout from these sensors triggered mechanical stimulation of the forearm, thus delivering physiologically relevant feedback information to the wearer. Furthermore, the relevant performance metrics of the system were tested for various variants, and the most appropriate set of parameters were selected for use in the clinical studies (such as minimal force threshold).

Figure 3: The custom liner with the integrated actuators, the fitted socket and the hand prosthesis with force sensors.

The sensory feedback system includes a "bubble-sensor" matrix, where silicone encapsulated gas bubble are connected via a thin silicone tube to pressure sensors or directly to pneumatic actuators. A pneumatic actuator is used to transfer sensations to the forearm skin from a hand prosthesis equipped with sensors.
Second, a system for open loop TENS was developed. In this case, the surface electrical stimulation was delivered independently of a robotic hand. To successfully control stimulation therapy sessions, a complete control software was developed: the psychophysical testing platform. The ‘Psychophysical testing platform’ is a dedicated software with the aim to provide time sensitive control over all aspects of the experimental setup. The setup was composed of surface electrodes applied to the stump. The computer controlled two screens, one for displaying information to the experimenter, and the second for displaying patient specific information.
The complete system, including the software platform and stimulator, were tested together and shown to work optimally for a wide range of operating conditions and therapy types.
Third, a system for TENS feedback during hand prosthesis use was developed. This system was composed of a hand with incorporated tensions sensors, connected to a central processing unit (a single board computer).

Figure 4: an overview of the components used in the system for non-invasive electrical stimulation, including the hand prosthesis: (1) the sEMG acquisition device, connected to the forearm via gel electrodes, (2) the prosthetic hand, (3) the single-board computer, and (4) the TENS stimulator, connected to the skin via gel electrodes.

The overall system was tested extensively on healthy individuals, to fine tune the various parameters and timing issues which arose from the complex interplay of several independent components exchanging data in real-time. The system was found to be robust, and the best parameters were selected for use in the clinical study.
Finally, initial preparatory steps for the future commercialisation of the system were carried out during the project’s timeframe. These included preliminary market research, as well as initial considerations considering CE marking and relevant regulations. Overall, clear steps have been defined for the future work needed regarding commercialisation.

WP3. Technologies for direct, invasive peripheral nerve sensory feedback

WP leader: ALU-FR

The objective of WP3 was to develop a platform applicable for invasive peripheral nerve sensory feedback studies on human subjects. The platform should be applied in patients suffering from phantom limb pain using either sensory feedback without or with an additional hand prosthesis.. The platform should include (1) intrafascicular nerve electrodes for chronic implantation, (2) a wearable electrical stimulator connected either to (3) a psychophysical testing platform and/or (4) a hand prosthesis.

Overview of main results
• Technologies to drive direct, nerve sensory feedback (TIME-4H, STIMEP and hand prosthesis) were adjusted, refined and successfully integrated with focus on long-term stability.
• A psychophysical testing platform was developed.
• The technical teams assisted to establish and refine clinical protocol and provided technical support to the clinical test teams during the surgeries and afterwards.
• Preparatory work for commercialization of the system was done.

Details of main results
The main goal of WP3 was to develop, refine and adjust the technical devices for the invasive clinical trials. The whole technical setup consists of four sub systems, (1) the transversal intrafascicular multichannel electrode (TIME), which is implanted in the median and/or ulnar nerve, (2) the stimulator STIMEP, which is capable of driving four electrodes in parallel, (3) a hand prosthesis equipped with sensors for haptic feedback and (4) a psychophysical platform to apply therapy and to obtain quantitative and qualitative measures of the effect of the therapy (Figure 5)
Partner ALU-FR has adjusted and refined the previous TIME-3H from the EU-FP7 project called TIME (sub chronic application) to end up in the current version, TIME-4H (chronic application). In depth analysis of the TIME-3H explants led to the following optimization steps:
First, the design and positioning of active sites and ground sites was changed, but the established outer dimensions of the electrode (width and thickness) remained within the previous version. Changes included the segmentation of the ground electrode to reduce intrinsic stress and delamination. Furthermore, the location of active stimulation sites on the substrate was shifted and additional adhesion promotion layers without access to the material-tissue-interface were included (layer setup was examined according to the ISO 10993 standard on cytotoxicity and passed) and tested with the partner UAB in the small animal model. Electrode sites proved to be more stable during continuous stimulation in vivo with adhesion layers compared to sites without. On the other hand, the packaging of the system was changed in some details, like the linkage between cable and connector or the identification tags. Furthermore, the restricted materials, which are in contact with the tissue after implantation of the devices, have been exchanged to non-restricted ones. These changes have also been integrated in the quality management system of the partner ALU-FR (fully ISO 13485 certified). Based on suggestion of US FDA guidance, ASTM F2503, IEC 62570 and for CE marking, the TIME 4H implants were investigated on MR compatibility by the company MR:comp. This non-clinical testing has demonstrated, that the TIME-4H implant is MR conditional in clinical relevant position and orientation according to the specification of ALU-FR. A patient with this device can be safely scanned in an MR system meeting the conditions specified in the manual. Essential validations for usage of the implants within invasive clinical trials, like washing and sterilization validation, were accomplished as well.
After design freeze the implants were constantly further improved in detailed parts (e.g. epoxy encapsulation, plasma treatment) to increase the long-term stability for the clinical trials (e.g. after analysis of the first round).
Both French partners, MXM-OBELIA and UM, refined the stimulator and stimulation paradigms according to the clinical requirements and to the performances of the considered system. The development, design and manufacturing of the miniaturized multichannel stimulator STIMEP included:
• 1 base board: BAISE_EP supporting power supplies, peripherals and insulated communication links and carrying
• 4 x 24xEP: independent 14 channels stimulator front-end, based on dedicated MXM-OBELIA KF24 ASIC component
• 1 M2S-FG484 ―off the shelf microcontroller board
• Design and manufacturing of the miniaturized STIMEP packaging
• All accessories (cables, switchboard etc.) have been specified according to these specifications as commercially available components
• Interconnect box has been designed and manufactured
Miniaturized devices were tested and integrated before final delivery according to the corresponding EU directives and international standards.
The partner UM designed and partially developed the controller-embedded software, as well as communication protocols with third parties (PLPP and HPC). UM has also co-designed with MXM-OBELIA, the global STIMEP architecture as well as the digital architecture of stimulation units. As the partner UM does not have an established quality management system special care was devoted to comply with the harmonized standard EN 62304 “Medical device software - Software life-cycle processes”.
The performance of different stimulation strategies for intraneural stimulation was assessed using TIME electrodes and the Stimn‘D stimulator (from the TIME project, delivered by UM and MXM-OBELIA) in experiments in rats at the UAB. The results showed that the inclusion of a 100 μs delay between the cathodic and the anodic phase of the stimulus allows to reduce charge requirements by around 30% without affecting stimulation selectivity. Further in vivo studies, were made to determine to which extent the stimulation charge might be decreased by delaying the discharge phase, depending on pulse duration and the way discharge is performed. The results demonstrated that delaying the discharge phase allows to gain more charge for shorter pulses. The difference in the gain between waveforms with passive and active discharge is lower.
In order to investigate the physiological bases and the most efficient therapeutical pattern of nerve stimulation for reduction of neuropathic pain, experiments on stimulus patterns have been performed in animal models of peripheral nerve lesions that cause hyperalgesia and pain.
Data generated by the small animal model testing was delivered to the partner UM for optimization of the STIMEP stimulation parameters.
EPFL was mainly responsible to integrate a sensor equipped hand prosthesis for the sensory feedback. Therefore, EPFL has developed encoding algorithms to transduce the readout of sensors embedded in the prosthetic hand in stimulation parameters. Moreover, EPFL has implemented decoding algorithms and the control of the prosthetic hand.
These algorithms have been ported on a small controller that can be powered with a battery (Odroid U3). Two systems can be used for acquiring sEMG signals which are Grapevine (Ripple) and TeleMyo (Noraxon). Moreover, two robotic hands were available: Azzurra (Prensilia) or the Robotic hand from Wessling Robotics. As planned, the control of the STIMEP, for driving the current injection in the intraneural electrodes, has been implemented. The chosen controller was able to virtually interface to every device wirelessly or with wires (USB or SPI connection).

Figure 5. ALU-FR developed, refined and adjusted the transverse intrafascicular multichannel electrode to the newest version TIME-4H, which was used during the clinical trial to interface the median and ulnar nerves electrically. The STIMEP stimulator, developed by UM (software) and MXM-OBELIA (hardware), was used for driving the TIME-4H. Moreover, EPFL developed the possibility to link the STIMEP with a sensor equipped hand prosthesis to realize sensory feedback.

The Psychophysical Testing Platform developed by the partner AAU is a computer based system developed to apply therapy to modulate painful and non-painful phantom limb sensations and to obtain quantitative and qualitative measures of the effect of the therapy. The Platform also assists to secure that the therapy is delivered and the data is collected in a systematic and time efficient way across clinical test sites.
Since the Psychophysical Testing Platform was intended to control the STIMEP, which drives electrical stimulation administered by the implanted TIME-4H devices, the software was therefore classified as IIb according to directive 93/42/EEC annex IX.
Function included to the platform were:
• Log in control of administrators, experimenters and experimental subjects
• Keeping track of the experimental phase of each subject and which questionnaires and test to present to
• Monitoring the subjects
• All questionnaires included in common protocol (D1.1) in Danish, English, Italian and French.
• Different images for left/right and arm/leg amputation.
• Save logs for conducted procedures and results of questionnaires.
• Facilities for planning and conducting stimulation using the STIMEP (iE) or InoMed (niE)
• Review functions
In accordance to prove the functionality of the all devices and systems, animal tests were performed on rats (UAB) and pigs (AAU). All the procedures used by the psychophysical testing platform (PSYP / USB), the hand prosthesis (HP / SPI) and SYNERGY (USB) software (testing software used by UM) were fully assessed using the whole EPIONE system (TIME-4H implanted on the sciatic nerve of the rat, STIMEP and final versions of software). Typical measurements with both EMG recordings and synchronization output, and evoked EMG on 3 muscles (PL, GM, TA) were performed. Safety embedded procedures was also successfully tested. During the use of STIMEP and third party software within the clinical trial, no critical bugs were reported. ALU-FR fabricated according to the quality management system requests more than 60 implantable TIME 4H devices. They are all listed in the system and traceable. MXM completed manufacturing (12 pieces), verification and validation of all STIMEP systems (10 systems in total) before dispatch to the partners. Each device and accessories were thoroughly tracked and traced to ensure traceability as requested by the quality management system. The WP3 teams provided technical assistance in two human implantations. A technical dossier is set up as documentation and prerequisite for further clinical trials and commercialization steps.

WP4. Clinical trials: Efficacy of non-invasive sensory feedback


The main objective of this work package was to assess efficacy of the non-invasive feedback to reduce phantom limb pain (PLP) using electrical or mechanical sensory feedback with or without the simultaneous operation of a hand prosthesis device.

Overview of main results
• Assist to establish and refine EPIONE clinical protocol
• Testing the non-invasive system/methods with the developed EPIONE clinical protocol
• Analysing and reporting clinical result of the non-invasive clinical therapies.
• Mechanical and electrical stimulation of referred sensation areas provided consistent positive non-painful evoked sensations, qualified for sensory input for therapy of phantom limb pain.
• Experience of phantom limb pain was highly individual, however, common features of the types of pain could be identified for groups of patients.
• Out of 23 subjects completing the therapy session we observed a consistent decrease of pain in 10 subjects as indicated by the VAS measure questionnaire during or after the therapy session, and a significant short-term relief of phantom limb pain has been observed in 12 subjects during the therapy session.
• Selective effective relief of specific components of pain associated with changes in the unnatural perception of the phantom limb associated with pain relief were reported.

Details of main results
Four types of non-invasive sensory feedback were used during a four week therapy session in the clinical tests electrical and mechanical feedback with or without the operation of a hand prosthesis device. The patient groups included upper limb amputees, lower limb amputees and different type of nerve injury patients experiencing phantom limb pain. Only the upper limb amputees worked with the handprosthesis device. A variety of clinical observations were reported that indicated a strong correlation between the stimulus delivered, characteristics of the referred sensation areas, external factors and the type of pain experienced. The outcome from the clinical trials was feed back to WP1 (Clinical trials governance) where a joint group analysis was performed. In WP4 the analysis and reports was instead done by a case-by-case analysis.

Stimulation strategy – utilizing the referred sensation areas
Before initiating the therapy phase we examined and mapped the referred sensation areas (RSA) for each subject. The RSAs were highly specific to each subject, depending on e.g. level of amputation, on the condition that led to amputation, and on the type, intensity, and duration of the stimulus. RSAs occur as nerve endings sprout following amputation connecting to existing skin sensors, representing a gate to the nervous system to provide sensory input to the brain. If the subject had a phantom map referred sensation area, this was then used as stimulation zone for the therapeutic stimulation. The phantom map referred sensation area was defined as an area of skin on the amputee that is if touched gives the amputee the sensation of touching part of the amputated limb.
In some subjects RSAs could be determined as larger areas on the skin, whereas in the other subjects RSAs were characterized as smaller spots. We found that the RSAs maps typically changed slightly over time with respect to in both location and type of sensation. Also, we found that a majority of the subjects would at first report not to experience any referred sensations, while this changed after some sessions.

Characteristics of the experienced phantom limb pain
Individual experiences of phantom limb pain (PLP) were reported, as perception of pain is highly individually and it may be affected by a great variety of factors. However, common features were noticed with subjects experiencing similar injury types (arm or leg amputees, as well as brachial plexus nerve damage) described the type and pattern/dynamics of PLP such as:
• Hard or partial clenched fist (upper limb amputees)
• Stretched / compressed (like small shoe, tight ring, or vice like clamping), or cramps in muscle of phantom limb (lower limb amputees)
• Short but very intense iron bar/needle piercing or current
• Constant continuous pain
• Continuous pain with relative fast and large variations within seconds/minutes
• Additional factors affecting dynamics of phantom limb pain: weather (warm, cold, pressure), stump load during physical activity, associated with use of prosthesis, intake of medicine on demand.
The naturally occurring dynamics of the experienced pain (i.e. the changes in pain over time), especially the frequent variations, posed additional demands when evaluation of evoked/induced sensations.

Short-term effects of therapy by surface electrical stimulation
The type of stimulus, location for delivery (i.e. the referred sensation area), and dosage proved to affect the type and intensity of phantom pain experienced. The following were noticed during repeated/randomized stimulus delivery:
• Relative long delayed occurrence of significant pain reduction. A sub-group of upper-limb amputee subjects initially had the feeling of a hard clenched fist. The ‘clenched fist’ opened after several session and were associated with a pain reduction by approximately 30-40%.
• Immediate pain modulation. A sub-group of subjects experienced immediate reduction of pain between 40 to 100%, with long-lasting effect, i.e. after stop of stimulus delivery, pain increased after a few minutes to the level experienced just before stimulus delivery.
• Short-medium delayed pain modulation. A sub-group of subjects experienced a delayed (seconds to minutes, dependent on pain level before stimulus delivery) pain reduction that lasted up to 3 hours.
• General minor reduction of pain during and after the therapy phase. A sub-group of subject not report any of the effects listed above, however mild reduction in pain level and less frequent pain episodes were reported during and after the therapy phase that correlated with an improvement in their general mood.

WP5. Clinical trials: Efficacy of direct, peripheral nerve sensory feedback

WP leader: UCSC
Partners involved: UCSC, CHUV, RN-Aalborg UH

The main objective of this work package is to assess efficacy of direct, peripheral nerve sensory feedback to reduce/suppress phantom limb pain (PLP) with or without the simultaneous operation of a hand prosthesis device.

Overview of main results
• We confirmed the possibility to use TIME electrodes for delivering intraneural stimulation and PLP treatment.
• Following the intraneural stimulation one of the two recruited patients experienced a statistically significant decrease of PLP intensity. In this patient the multimodal analysis for brain plasticity showed significant changes following the treatment, mainly on brain regions corresponding to the amputated side.
• Recommended procedures for surgical TIME electrodes implantation and for the intraneural stimulation process were provided.
• Some important lessons were learned regarding difficulties on patient recruitment and ethical and legal approval in this field.

Details of main results
By the end of the project it was only possible to recruit two out six planned patients. The two patients were recruited by UCSC. Despite strong efforts made, CHUV and RN-Aalborg UH were not able to find patients for the invasive trial. RN-Aalborg screened more that 30 subjects but nobody was recruited due to the inclusion/exclusion criteria. CHUV, due to the late approval from Swiss Medic, started the recruitment with delay and had the possibility to screen few patients. Also in this case the patients did not fulfil the inclusion/exclusion criteria.

PLP treatment
The first patient was a 37-year-old male from Ecuador with a traumatic transradial (proximal third of the forearm) amputation of the left arm occurred about 3 years ago. The second patient was an Italian 54-year-old female with a traumatic very proximal trans-radial amputation of the left arm occurred about 1 year and 6 months ago.
We confirmed the possibility to use TIME electrodes for delivering intraneural stimulation for PLP treatment.
Following the intraneural stimulation one of the two patients experienced a statistically significant decrease of PLP intensity (Figure 6 and Figure 7) measured with VAS (in the 24 hours) and NPSI. In this patient the simple intraneural stimulation was able to reduce the pain, while in the other patient PLP remained unchanged so far (the trial with this patient is presently on-going) also after prolonged session of continuous intraneural stimulation (VAS and NPSI score not statistically significant). Due to the low number of patients it is impossible to identify factors (e.g. like time from amputation, level of amputation, degree of functionality of the residual stump) linked to a possible success or failure of the treatment.

Figure 6. Graphic showing the comparison of VAS score before the implant and during the treatment.

Figure 7. Graphic showing the comparison of NPSI score before the implant and during the treatment.

Figure 8. MEP maps. A progressive enlargement of the excitable area of forearm muscles on the hemisphere contralateral to the amputation is evident during the trial.

Brain plasticity
The multimodal analysis for brain plasticity assessment has been completed in the fist patient, and it is being performed in the second patient whose trial is on going (in this case during the first five weeks of treatment it was possible to perform only the baseline evaluation). In this first patient TMS pre-therapy showed a slight abnormal inter-hemispheric asymmetry of motor cortex topography, resulting in a smaller area of representation of muscles governing the amputated limb compared to the area for the intact limb. Following therapy, cortical maps showed a partial reversal of this asymmetry because of an enlargement of the excitable area on the right hemisphere, contralateral to the stump, leading towards a more symmetrical muscle representation in the two hemispheres, as in normal subjects (Figure 8).
EEG analysis showed a progressive increase of alpha 1 power density during the trial, and a consensual decrease of delta band power density, mainly located in the parietal, occipital, temporal and central areas of the brain. The network analysis showed a statistically significant increase of the Path Length index in the sensory-motor network involving both hemispheres with a preferential increase on the right hemisphere (the contralateral to the amputation in this case). Functional MRI results showed a progressive reduction of motor areas needed to perform a task of phantom hand movement during the trial, indicating the presence of a motor learning process involving the amputated arm. On the other side structural MRI was not able to show significant changes of corticospinal tract and thickness of S1-M1 cortex suggesting that in our experiment the brain changes globally observed in our patients were mainly due to functional plastic phenomenon acting at synaptic level.

Lessons learned
The process of ethical and especially legal approval of the invasive part of EPIONE project was complicated. First of all, every European nation has is own legislation, meaning similar but not equal documentation had to be prepared by clinical partner in the project and no recruitment was possible until the study was legally approved some precious time was lost. Moreover it proved more difficult than expected to recruit patients for the invasive clinical trial mainly because of the combination of some issues: 1) Few potential subjects experience strong PLP and the numbers available in the literature may be overestimated; 2) The defined inclusion criteria (e.g. site and level of injury, etc to allow space for the implants), made it difficult to include most of the interested patients, so future studies focused on PLP treatment should include also lower limb amputees; 3) The temporary effect (if any) of the treatment, the hypothetical risk related to the trial (in particular because the system was percutaneous and therefore prone to mechanical damage and infection) and the need to perform two surgical interventions (the first to implant and the second to explant the TIME electrodes); a completely implantable system could have been more acceptable. These experiences will be important to speed up the approval process in eventual future studies.

WP6. Exploitation and dissemination

WP leader: Novosense AB
Partners involved: All partners.

The RTD developments within EPIONE was expected to produce several important knowledge and technology-based results. Implementation of active strategies for the exploitation and dissemination of the results is fundamental for bringing the results and developments to the market in the future. The objectives of
the exploitation has therefore been to develop a sound set of business models around the EPIONE technologies being developed. The objective of the dissemination is to develop a plan to facilitate the translation of the research results to the general public and other stakeholders.

Overview of main results and finalised tasks
• Dissemination plan for patient recruitment established
• Established dissemination plan for communication with the scientific community and industry
• A market analysis
• A competitor landscape analysis
• A IPR strategy analysis report for project partners
• Developing a concrete business model base on EPIONE results

Details of main results


Plan for pt recruitment
The following channels were identified for communicating to the target group.
• EPIONE website
• Partner local websites
• Press release
• News spots: Television
• News spots: newspapers
• Personal communication
• News spots: Radio
• News spots: magazines
• Conference
• Brochure
• Newsletter
• Social media
• Clinical trials registration
An plan was set up for utilising these channels and the general planning was to;
• Step 1 - Press releases
• Step 2 - Information on clinical trials and opportunities on EPIONE website
o for volunteers
o for medical and scientific community
• Step 3 – Partner level dissemination
• Step 4 – Evaluate
• Step 5 – Revise if necessary and iterate step 2 and 5 again.
More details on partners efforts for PT-recruitment has been reported in the earlier period reports.

Dissemination plan
The first version of the EPIONE dissemination plan was reported M14 and it has thereafter periodically been updated. The overall plan is shown in table below.

No. Title Description of Activities Main leader Date Place Audience Countries addressed
1. Home page Raise awareness AAU Oct 2013 Internet www all
2. Logo Raise awareness AAU Oct 2013 Internet www all
3. Advertisement for clinical trial Raise awareness of the trial and potential treatment Clinical trials team Pending ethical committee and PO approval Internet, Print, Mail Clinicians and prospective subjects
4. Press release Inform the public UAB July 2014 Public media (newspaper, radio) General public Spain
5. Meeting presentation Communication to the FENS meeting UAB July 2014 Milano, IT Neuroscientists EU
6. Hospital and University Journals/Newsletters Raise awareness CHUV Jan 2015 Switzerland University and patients Switzerland
7. CHUV Intranet/Department website Raise awareness CHUV Jan 2015 Intranet CHUV employees Switzerland
8. EPIONE - workshop at international conference
AAU/entire consortium year 3 and/or year 4 Do be determined Scientific community all
Table 4. Dissemination planning

The last joint dissemination task, “8. EPIONE - Workshop...” was later decided to be to set up a special session at International Conference on NeuroRehabilitation, ICNR2016 and at the RehabWeek in London 2017.


Competitive landscape
The work in regards to mapping the competitive landscape led to the identification of 90 organisations that very diverse in nature, including research companies, larger medical companies and research universities. These were then analysed according to the business model configurations that they apply. For various reasons, including bankruptcy, M&A’s and secrecy, not all of these could be mapped in practice. This left a total sample of 80 companies.
Currently, the competition regarding Non-Invasive Technologies solutions for reducing PLP is relatively limited. Most of the solutions offered today are still at in a research and development phase. But nonetheless, given the positive impact of these technologies in reducing PLP (as indicated by pilot tests conducted), an increasing interest by various research centres to enter into this emerging industry, which, evidently was identified. This indicates a future potential growth in this industry. Because of the early stage of industry, the review does not always provide reliable data on the most probable business models and as such the next phase in this regard will take its outset in the understanding of the medico-tech sector.
Competition within the invasive technology segment varies depending on the market. In the USA, competition is strong, and mainly lead by Medtronic and the major hospitals, who offer various pain treatment options. This is partly because of the liberal healthcare system, which compared to Canada and most of Europe, is not financed by public funds. The general competition is also very much dependent upon whether or not the clinics offer treatment using their own product developments or health care technologies developed elsewhere. Access to these markets is regulated through national health authorities, who approve/reject use of new health technologies.
Strong competition exists within Orthopaedics Prosthetics, where the two largest companies hold 75% of the total market share. Products in this category have been on the market for several years. With new innovative products, it could be possible to gain market access as there are many companies with similar products today.
Currently all pharmaceutical approaches in the medical segment will be offered as first line of therapy for reducing PLP, and may thus not be considered direct competitors for the EPIONE project technologies. The medication is generally inexpensive compared to the other types of treatment, as they are non-invasive and the prices of these are unlikely to increase in the future. However, it has been difficult to establish any proven substantial clinical effect of these on PLP, and thus EPIONE may be expected to acquire market share from these treatment options.

Identified EPIONE IPRs
The EPIONE exploitables was identified early in the project and an analysis regarding strategy, owner, etc. was delivered at month 16, see Table 10.2. The list of exploitable has been update a few time during the duration of project.

Business models
The work with the EPIONE business models included several subtasks such as identifying exploitable EPIONE outcomes and stake holders, the making of market analysis, investigating competitors, IPR strategies and product cost analysis. All these activities have then lead to the final development of the EPIONE business models. In short the following strategies was suggested.
• The clinic-based business model: Here the technologies would be sold directly to PLP patients through direct contact in clinics. This disintermediates the existing structures, where PLP patients mainly are treated through university hospitals. Such clinics take on the business model of being trusted advisors. They can either be core-focused (i.e. only on PLP), or they can have a diversified more full-service orientation towards pain treatment.
• The prosthetics-based business model: In essence, this business model differentiates itself from the clinic-based business model by being a product-based, with three possible routes. (1) a long-tail scenario for selling a wider scope of “non-hit” products in low quantity, (2) a cell-phone scenario offering different plans in relation to a product featuring a range of prices depending on varying levels of usage, or (3) reverse bait and hook offering a low-margin product at low or no cost to encourage sales of the initial higher-margin product
• The data-driven business model: A third opportunity of a business model configuration was identified as the ‘data-driven business model’. Being a data-based model the scalability is much higher than man-hour and product-based business models, making it easier to achieve true scalability. This business model would entail developing the software platform where the EPIONE technologies would provide the basis of a trusted advisor position and a trusted product/service leadership strategy.

Potential Impact:
The potential impact of EPIONE has been analysed in relation to 1) Science and Technology, 2) Economy and Society, 4) The political landscape, 4) Exploitation and IPR and 5) Dissemination

Science and Technology
The originally defined expected impact of the HEALTH 2013.2.2.1-5 topic was: ‘’(a) Successful projects are expected to deepen our knowledge of how pain is generated, propagated and quenched, (b) work towards the identification of more effective diagnostic and/or treatment approaches and (c) help translate pre-clinical and clinical results into solutions for the patients’’

The EPIONE consortium provided input to the expected impact of the topic through the following.
a) Clinical trials outcomes. We investigated if generation of PLP may be explained by changes in the cortical map that follows amputation, and whether PLP may be quenched by restoring the cortical map through sensory feedback therapy. We compared two main routes for providing a more long-term or permanent solution for the amputees, i.e. a non-invasive sensory feedback solution that could be integrated with the operation of a hand prosthesis and an invasive sensory feedback solution that in the future may offer the ultimate, permanent, invisible and cosmetically acceptable interface.
A total of 31 volunteer subjects were enrolled in the studies, 23 subjects started the therapy phase and 20 subjects completed the full study (the remaining withdrew during different phases of the study). Initially our focus was on upper limb amputees only, however, it was difficult to recruit subjects across Europe and the USA. The inclusion criteria were therefore widened to also include and lower limb amputees and subjects with complex regional pain syndrome that experience phantom limb symptoms in the trials where this was possible.
The group analysis from the trial indicated that there was, on average, a significant reduction in the phantom limb pain. It was not possible to do a group analysis on the effect of the therapy on the cortical organization due to the large heterogeneity of the included subjects.
We summarized our clinical experiences in a guideline for delivering sensory feedback.
b) Technological developments. Before EPIONE was initiated, the specific technologies were only available as experimental units and have only been tested on very few subjects. Implementation of EPIONE allowed a translation of experimental work and results previously funded by EU ( into the clinic for further validation. The validation is an important step to move the technologies into the medical market.
• In case of the invasive technologies, the research conducted within this project showed that thin-film electrodes embedded in polyimide can be safely and reliably implanted for stimulation purposes in a human peripheral nerve. Intuitive sensory feedback can be restored with this type of intrafascicular neural multichannel electrodes (TIME 4H). The results that we achieved regarding the TIME electrode performance in two human volunteer subjects and use for phantom limb pain treatment are very important for applications of these electrodes in the closely related field of sensorised robotic hand prosthesis in amputees. To transfer into a permanent solution and thereby make the system commercially available will be dependent on finding a solution to connect single TIME 4H electrodes via a detachable plug system to implantable electrical stimulators. Further clinical validation and assessment would be needed for the invasive system, where it showed only possible to recruit 2 of planned 8 participants.
• In case of the non-invasive sensory feedback we developed novel non-invasive technologies (electrical or mechanical systems with or without hand prosthesis integration) for delivering sensory. A dedicated software platform was developed to implement the clinical protocol and gather the clinical data. In addition dedicated software was developed to extract and analyse the common clinical data. In case of the non-invasive systems, system miniaturisation and human factor engineering are the main requirements to effectively translate the non-invasive technologies developed during the EPIONE project into commercial products. The focus of these efforts would be to transform research prototypes into standalone therapeutic devices, easily useable by the patient or medical personnel. Preliminary steps for the preparation of commercialisation were carried out during this project, which could be leveraged by a party willing to pursue the successive steps.

Economy and society
The originally defined overarching objective of the call was to ’’(a) improve the health of the EU citizens and (b) increase the competitiveness and boosting the innovative capacity of European health-related industries while addressing global health issues’’.

We believe that EPIONE provided direct input to this objective through the following.
a) There are currently more than 1.6 million persons in the US and more than 400.000 persons within Europe that are living with the loss of a limb (2005 numbers) and many of these are living with phantom limb pain. A potential means to treat PLP to those suffering from its effect was demonstrated. For both upper and lower extremity who benefitted from the therapy, the therapy represents a means to reduce the impact of PLP on their lives through the reduction or potentially elimination of the use of drugs prescribed to control pain. In the case of the lower extremity subjects who benefited from the therapy, the therapy represents an opportunity to use their prosthetic legs and regain mobility and freedom that they lost as a result of amputation and PLP. The successful deployment of the therapy could represent an improved quality of life to these subjects, which could result in the decrease of time lost from work due to PLP, time spent in out-patient care to treat pain, and the reduction in the prescription of prescription analgesics which come with the potential side-effect of addiction and substance abuse.
b) In case of the invasive system, the technical system opens new therapy options for subjects with amputation trauma to increase their quality of life in long-term perspectives. Sensory feedback is a first step for better embodiment of prostheses and more frequent use.
c) In case of the non-invasive system, the various non-invasive sensory feedback systems developed during the project have great potential for socio-economic impact. Indeed, the developed systems could serve as the basis for commercial devices which may help alleviate phantom limb pain symptoms with minimal invasiveness, limited side effects and relatively low cost. This contrasts with approaches requiring surgical intervention or based on pharmaceutical. Preparatory work for the commercialisation of such devices was carried out during the project, facilitating future impact of these non-invasive technologies.
d) Even though that EPIONE provided a proof of concept, the developed systems are not yet mature enough for immediate commercialization. However, identified short term effects include; 1) Opportunities for younger scientist to learn and network with other European scientist, 2) Established scientific cooperation between several European and one American scientist group. 3) Increased knowledge of scientists regarding medical regulations

The political landscape
Within the political landscape as defined in Horizon 2020 and the Europe 2020 strategy, one of the seven identified flagships is ‘The Innovative Union’ [weblink 1, weblink 2, weblink 3], where there is a focus on fostering new ideas to tackle societal challenges and ensuring that EU investments can be properly exploited. One of the identified challenges is ‘Health – demographic change and well being’ [weblink 2]. In particular the Innovative Union should ‘’respond to the major health-related socio-economic and societal challenges in view of the new orientations given by the EUROPE 2020 strategy’’

We believe that EPIONE provided direct input to this objective through the following.
• Science. A better understanding of which factors are affecting the degree of phantom limb pain and pain management
• Science. Which sensory feedback type is efficient in quenching PLP
• Technology. Dedicated technologies to deliver different types of sensory feedback developed, delivered and tested.
• Technology. Preparation of technologies closer to the medical market
• Economy and society. Assist to maintain knowledge and expertise within EU (the brain drain problem)
• Economy and society. Knowledge transfer from academia to industry
• Economy and society. Better positioning of SMEs within the European health-related industry (partners MXM and NS)
• Economy and society. Training of younger researchers
• Economy and society. Opportunity for younger researchers to establish a European/US network

In EPIONE we carried out clinical trials in 4 European Countries (Denmark, Sweden, Switzerland, Italy) and in the US. The realization of the project has shown, that the project duration of four years is a strict timeframe to refine technologies, obtain approval and implement the clinical trials. Since medical device directives have always to be transferred into national law with national legal bodies, it is therefore a highly labour intensive and challenging task to plan clinical trials in several European countries. Even though non-approval studies are now separately addressed in the new European Medical Device Regulation (EC 2017/745) for all European companies, detail specification what has to be done beyond the “essential requirements” has to be delegated to the member states and their specific national laws.

Exploitation and IPR
EPIONE has resulted in two filed patents and two registered software applications. The exploitation framework was analysed during the project and dedicated business models for the EPIONE specific technologies were devised. Although the different non-invasive technologies need final development and validation this is key knowledge in taking the final step to bring the technologies to the medical market. Due to the low number of subjects recruited for the invasive technologies, this system must undergo further clinical validation.
The EPIONE work has resulted in 31 peer reviewed articles/conference proceedings, with a number of pending and planned publications. The consortium has disseminated their work at more than 160 events during the project life-time, including several tv/radio/press coverage for the general public, and two organized special-sessions at international conferences.

List of Websites: