The present invention relates to a medical device. The invention is of particular relevance to implantable devices, for example those interfacing with biological tissue such as the brain and/or for the delivery of therapeutic substances.

It is estimated that neurological disorders, such as Parkinson’s disease, Alzheimer’s disease and depression affect hundreds of millions of people worldwide. However, despite such a high need, the success rate for existing treatments of central nervous system disorders is very low.

It is likely that there are two main reasons for this low success rate. The first is an incomplete understanding of the brain, particularly at the macroscopic level. The second is the difficulty in achieving a constant targeted delivery of therapeutics due to the blood-brain barrier (BBB) and the blood-cerebrospinal fluid barrier (Wolak et al. 2013). It is estimated that the BBB blocks approximately 98% of small-molecule drugs and nearly 100% of large-molecule drugs (Zhou et al. 2013).

Various attempts have been made to develop a delivery method which is able to overcome the BBB. One proposed solution is electrophoretic drug delivery which allows a dry delivery by letting only the drug molecules pass through a membrane into the brain. However, this method is limited to certain applications as the drug needs to be small and charged; it is also difficult to deliver larger amounts of the drug (Proctor et al. 2018).

Another proposed solution is “convection-enhanced delivery” (CED), which is a technique that harness increased flow of fluid to increase transport of large molecules and drugs throughout a tissue. CED works through applying a local pressure differential to drive fluid flow through the tissue so that convective forces dominate over diffusive transport. This allows drugs to bypass the BBB.

However, CED has a number of drawbacks. The needle used is typically in the range of several hundred micrometres to a few millimetres in diameter, potentially triggering a foreign body response and making it prone to reflux. Furthermore, the pressure applied may damage the brain and cause edema. The present invention aims to solve one or more of the above problems.

Aspects of the present invention aim to provide devices that can mitigate or even eliminate foreign body response.

Aspects of the present invention aim to provide devices that have mechanical properties matching the brain.

At their most general, aspects of the present invention provide minimally-invasive devices which can deliver a drug to an area of the body, such as the brain, at a cellular scale, preferably using mainly diffusion as a transport mechanism.

A first aspect of the present invention provides an implantable medical device having: a flexible substrate; a delivery conduit formed in the flexible substrate and having an inlet and an outlet located at a proximal end of the device, wherein the delivery conduit passes from the inlet to the outlet via a loop at a distal end of the device; and a delivery hole formed in the delivery conduit allowing passage of fluid from the conduit to the exterior of the substrate.

The delivery conduit can thus form an enclosed fluid passageway from the inlet to the outlet with fluid only able to exit the conduit through the delivery hole. This circulation-type arrangement can allow for diffusive delivery (or at least a predominantly diffusive delivery) of a medicament, by the medicament diffusing from a fluid in the delivery conduit through the delivery hole, rather than a convective delivery which would result from one or more open- ended delivery conduits through which a medicament is delivered.

In particular, the arrangement of the device may allow for circulation of a fluid containing a medicament to and through the implantation area, with relevant molecules being able to diffuse into the area around the delivery hole, but with only a fraction of the fluid supplied to the delivery conduit actually exiting the conduit. This can be contrasted to the alternative of forced delivery of a quantity of fluid to the distal end of a more standard “injection-type” delivery conduit in which all fluid delivered to the distal end is forced into the area of the body surrounding that end.

The delivery hole is preferably formed at the distal end of the device. The device may have multiple delivery conduits each formed into loops. These conduits may be in fluid communication with each other (for example sharing a common inlet and a common outlet between two or more of the conduits), or may have separate inlets and outlets, or a combination of these alternatives.

The device can be arranged to deliver any type of drug or medicament through the delivery hole.

The device is preferably sized and configured to minimize or eliminate foreign body response. This may include the selection of materials that the device is manufactured from. For example, sizing the device at the cellular scale has been shown to mitigate foreign body response and allow for long-term implantation.

Preferably not only the substrate but the entire device, or at least the implantable/distal end of the device (for example the portion which is intended to contact the target tissue) is flexible, which preferably includes the sensors. The flexibility can also help to reduce foreign body response.

In particular embodiments the device, or the distal end thereof, may have a bend radius of no more than about 2 mm, preferably no more than 1.5 mm and more preferably no more than 1 mm. Bend radius, which is measured to the inside curvature, is the minimum radius that a component (in this case the device) can be bent in at least one direction without damaging it. The bend radius as defined here refers to elastic deformation as opposed to plastic deformation such that a device bent under an applied force to a radius greater than the minimum bend radius would return at least part way to its original shape with the removal of the applied force. In other words, in these embodiments, the device of this aspect can be bent to an inside curvature of 2mm, for example by rolling when the device is being arranged for insertion into a patient, and subsequently deployed (e.g. unrolled) to an expanded, less bent configuration (e.g. a substantially planar configuration) and still function exactly as it did prior to bending. The device may further comprise one or more sensors. One or more of the sensors may be formed on an outer surface of the flexible substrate. Alternatively or additionally, one or more sensors may be formed in the delivery conduit.

The sensor may comprise one or more electrodes. The sensor may be positioned proximate the delivery hole.

The sensor may be arranged to detect the delivery of a medicament from the device and/or to detect a condition of the patient, for example by monitoring brain activity.

The sensor(s) in the delivery conduit may be arranged to monitor flow rate or pressure.

The sensor(s) may be arranged to detect the presence of a biochemical.

There may be a plurality of sensors, which may be arranged for the same purpose of for different purposes. Sensors may be arranged to cooperate, for example so that a difference between the signal from two (or more) sensors provides relevant information. The plurality of sensors may be arranged in an array which may include a regular spacing of sensors and/or location of sensors at known distances from each other and/or from the delivery hole.

The sensor(s) may be connected to the proximal end of the device by one or more conductors.

The device may have a plurality of delivery holes each allowing passage of fluid from the conduit to an exterior of the substrate. A plurality of holes may be provided to allow passage of fluid and molecules from the delivery conduit to the exterior of the substrate (for example into the patient’s body) at a variety of different locations, or to increase the available contact area through which transfer of fluid and molecules can take place.

In certain embodiments, a plurality of said delivery holes are arranged proximate each other.

In certain embodiments the delivery holes are arranged in a plurality of groups, the holes in each group being closer to one another than to holes in a different group. The delivery hole or delivery holes are preferably micron sized. More preferably the delivery hole or delivery holes have a diameter between 1 and 50 pm, preferably between 5 and 10 pm. Where there are a plurality of delivery holes these may be of identical sizes, but need not be.

One or more of the delivery holes may have a semi-permeable membrane which permits passage of molecules below a certain size through the delivery hole whilst preventing passage of molecules above said certain size through the delivery hole. This may allow, for example, relatively small medicament molecules to pass from the delivery conduit through the delivery hole to the patient, whilst preventing larger molecules (for example proteins) from passing in the opposite direction. In other embodiments the holes are not covered by any kind of membrane.

The external diameter of the delivery conduit is preferably between 1 and 500 pm, more preferably between 5 and 100 pm and more preferably between 5 and 25 pm.

The device is preferably configured for implantation through an incision of less than 6 mm diameter. This can enable the device to be used in minimally disruptive surgery. The device may be collapsible/foldable/rollable in order to fit through the incision.

Non-exclusive specific applications of the device of this aspect include delivery of neurotransmitters to study brain behaviour and the treatment of very localised diseases such as epilepsy and Parkinson’s disease.

A second aspect of the present invention provides a system for delivery of a medicament to a patient, the system including an implantable medical device according to the above first aspect, including some, all or none of the optional and preferred features of that aspect, and a controller arranged to control the flow of fluid through the delivery conduit of the implantable device.

The system may further include a pump in fluid communication with the delivery conduit, wherein the controller is arranged to control the pump. The pump may be any known pump suitable for this purpose, such as an electro-osmotic pump or a peristaltic pump.

The controller is preferably arranged to control the flow of fluid through the delivery conduit such that the medicament exits the delivery conduit through the delivery hole predominantly by diffusion rather than convection. The system may further include a supply reservoir from which fluid is introduced to the inlet of the delivery conduit and a waste reservoir in which fluid exiting the outlet of the delivery conduit is stored.

In certain embodiment the system has a plurality of supply reservoirs. The controller may then be arranged to control the supply reservoir from which fluid is introduced to the inlet of the delivery conduit. For example, the supply reservoirs may contain different medicaments and the controller arranged to select the medicament to be delivered.

Alternatively or additionally, the supply reservoirs may contain different concentrations of the same medicament. In this arrangement the controller can select the concentration of medicament to be supplied to the delivery conduit.

Alternatively or additionally the supply reservoirs may include a reservoir containing a concentrated solution of a medicament and a transport solution and the controller arranged to control the relative proportions of fluid from each reservoir in order to control the concentration of medicament in the fluid introduced to the inlet.

Alternatively or additionally, one of the supply reservoirs may contain a neutral or a cleaning solution to allow the delivery tube to be flushed or cleaned before and/or after the delivery of medicaments.

The implantable device may include a sensor and the controller may be arranged to receive a signal from the sensor and to control the flow of fluid through the delivery conduit based on the signal received from the sensor.

A third aspect of the invention provides a controller for an implantable medical device, the controller including a processor and a memory storing instructions for the processor, the controller being configured to: control a pump to deliver medicament to a patient through the implantable device such that the medicament circulates in the implantable device and exits the implantable device predominantly by diffusion rather than convection.

In order to cause the medicament to circulate and exit predominantly by diffusion, the controller may receive information, for example from sensors, about the flow rates at the inlet and the outlet of the device. In a diffusion regime, the flow rates should be equal or approximately equal. In other embodiments, the pressure at the inlet and/or outlet could be measured (instead of or in addition to the flow rates); for a diffusion regime the pressure should preferably be the same as, or almost the same as, the pressure inside the target tissue. In other embodiments, data from sensors implanted in the delivery zone could be used by the controller to determine and/or monitor the actually delivery of the medicament to the patient.

Preferably the controller is further arranged to: receive information from sensors on the implantable device; and adjust the control of the pump based on the information received from the sensors.

The controller may be used with an implantable medical device according to the above first aspect, including some, all or none of the optional and preferred features of that aspect, but need not be.

A fourth aspect of the present invention provides a method of delivering a medicament to a patient, the method including the steps of: implanting a delivery device having a flexible conduit configured allow for the passage of fluid from the exterior of the patient to an implantation site and for return of the fluid from the implantation site and having a delivery hole for delivery of a medicament in the fluid; and delivering the medicament through the flexible conduit such that the medicament exits through the delivery hole predominantly by diffusion rather than convection.

The method may further include sensing the results of the delivery using a sensor positioned proximate the implantation site and adjusting the delivery of the medicament based on the results from the sensor. The sensor may be part of the delivery device.

The medicament may be delivered to multiple sites through a single conduit.

Further aspects of the present invention provide a method of treating a brain disorder and a controller for an implantable medical device which is configured to perform that method.

The method of this aspect includes the steps of: receiving electrophysiology data from one or more sensors located in the brain of a patient; identifying, from the received data, a transition or change in the data which is indicative of a transition from a normal state to an abnormal state; and, if a transition is identified, controlling the delivery of a medicament to the brain of the patient through an implantable device located in the brain of the patient.

A particular example of the brain disorder is epileptic seizures. The method may use the electrophysiological data to identify an epileptic seizure or pre-seizure conditions and distinguish these from normal brain activity or to identify patterns in the data indicative of transition from a normal state to a seizure state.

The method may also including controlling which drug is delivered and the concentration of the drug delivered where the configuration of the device provides for such possibilities.

The method may include delivering a drug when an abnormal condition is identified in order to treat the condition. Alternatively or additionally the method may include delivering a drug when a transition from a normal state is identified, which may allow preventative treatment to avoid the transition to the abnormal state, or to delay or ameliorate the effects of the transition, or to lower the severity of the abnormal state.

The control of the delivery of the drug may include controlling a pump connected to implantable device. This can allow the control of both whether a drug solution is being delivered as well as the flow rate of that drug solution.

The method may also include the steps of receiving further electrophysiological data from the sensors and using that further data to identify when the patient returns to a normal condition, or when the patient’s condition stabilises. The method may include the step of adjusting the delivery of the drug accordingly, for example by stopping it.

Unless indicated otherwise, any of the features (including the optional or preferred features) described in relation to one of the above aspects are equally applicable in combination with the devices and systems of any of the other above-described aspects.

According to a further aspect of the invention, there is provided a method of treating a human or animal body, the method comprising implanting a medical device or bioelectric implant according to any of the variations of the method of the above aspect. The invention is described below, by way of example, with reference to the accompanying figures in which:

Figure 1 is a schematic of a device according to an embodiment of the present invention as modelled by a computational simulation;

Figure 2 shows the results of time-dependent simulation of a device according to an embodiment of the present invention using the model described in relation to Figure 1;

Figure 3 is a series of graphs showing comparisons of various profiles along an axial distance from the hole of the device simulated using the model descried in relation to Figure 1 at different flow rates;

Figure 4 is a series of plots showing how the amount of drug delivered varies in the model described in relation to Figure 1 with changing parameters;

Figures 5a and 5b are plots showing the effect of changes in hole geometry in the model described in relation to Figure 1;

Figures 5c and 5d are plots showing the spatial distribution of delivered drug after 1 second and 60 seconds respectively from a tube with three holes using the model described in relation to Figure 1;

Figures 6a and 6b are illustrations of a device according to an embodiment of the present invention when implanted in the brain;

Figure 7 is an illustration of a system according to an embodiment of the present invention;

Figure 8 is an illustration of another system according to an embodiment of the present invention;

Figure 9 shows the steps in a method of manufacturing a device according to an embodiment of the present invention; Figure 10 shows the step in an alternative method of manufacturing a device according to an embodiment of the present invention; and

Figure 11 shows a flow chart for a control system algorithm according to an embodiment of the present invention.

To analyse the operation of a device according to an embodiment of the present invention, a model of a U-shaped flexible microtube was designed with both inlet and outlet outside the brain and a small hole provided in the middle of the tube to allow diffusion of a drug circulating in the tube into the extracellular space. Figure 1 shows a schematic of this model.

A computational simulation of the model has been conducted by the inventors in order to understand aspects of the operation of the device. In particular, the simulation was used to: a) observe how the drug will be delivered by modelling and observing its concentration in the brain over time; b) determine whether it is possible to minimize the pressure applied on the brain in order to achieve a main diffusion-driven delivery; and c) observe the impact on the delivery process of using different drugs or different sizes or numbers of holes.

The simulation was performed using COMSOL Multiphysics®. All the simulations were timedependent and performed in 2D. The choice of 2D was to simplify the simulation whilst allowing a good understanding of the transport mechanisms involve and to observe the influence of the various parameters.

Mathematical Model

The brain is very densely packed with cells, whether considering grey matter or white matter. The transport of molecules occurs mainly in the space in between the cells, termed the extracellular space (ECS). This space accounts for approximately 20% of the total brain volume and can be described as a foam in which the cells correspond to the gaseous phase and the ECS represents the water phase.

Given that the ECS is orders of magnitude lower than the distance travelled by the drug, the brain can be treated as a porous media and Darcy’s law is thus applied: p'tJ. u = 0

In which a is the porosity of the brain, p is the density, u is the velocity vector, p is the pressure, p is the dynamic viscosity and K is the permeability.

The drug concentration over time is controlled by a modified diffusion equation (Sykova et al.):

The first term on the left represents how the concentration C change in time at a given location. The second term is the diffusion term with D* = I) / ² . The free diffusion coefficient D is reduced by the parameter A called tortuosity, which takes into account the geometry, the deadspaces, the obstructions and the binding happening inside the ECS. The third term of the equation is the introduction term, which is made zero in the present simulation as the introduction is included in the boundary conditions of the simulation. The fourth term represents the convection, driven by a corrected velocity vector and the last term incorporates loss, clearance and uptake. In the model, the diffusing molecules can be permanently removed from the ECS through the BBB, by entering a cell, by binding to receptors or by enzymatic degradation. These processes are typically proportional to the concentration and thus /(C) = keiim- C with k _eUm being the elimination rate.

These equations describe two transport mechanisms, which are diffusion and convection. Diffusion is driven by a concentration gradient and convection is driven by a pressure gradient.

Model Geometry

The optimal width of the tube was chosen to be 10pm which was considered to be sufficiently small to reduce the device footprint while still allowing a significant quantity of drug to be delivered. The distance from the tube inlet to the U portion and from the U portion to the outlet was set to 1cm, resulting in a total tube length of 2cm.

The size of the hole was set to 10pm, but the impact of smaller (5 pm) and larger (15 pm) holes was also studied. The effect of having several holes was assessed by implementing tubes with 2 and 3 holes. In the case of 2 holes, these were placed at +45° and -45° around the middle horizontal axis. In the case of 3 holes, there were placed at +90°, 0° and -90° around the same axis.

In order to avoid any influence of border effects of the simulation model, the brain was defined as a larger surface area than the striking distance of any drug in the timescale studied and thus a circle of 20mm radius was chosen.

Model Parameters and Assumptions

The inside of the tube was assumed to be filled with water and the brain tissue was treated as a porous media. Other material parameters needed for the simulation were taken from Zhan et al. and are shown in Table 1 below. Parameters set by the user or dependent on the drug being delivered are shown in Table 2 and described further below.

Table 1: Parameters for the tube and the brain tissue

Table 2: Parameters for the simulation

Flow rate: The flow rate used in the study was 3nL/min. The impact of different flow rates was assessed for rates ranging from InL/min to 7nL/min.

Pressure: The pressure in the brain, also called the intracranial pressure (ICP), varies a lot. It depends on several aspects such as the age, body posture and clinical condition and of course is different for humans and other animals. For this study an averaged ICP found in rats of 536Pa was chosen. However, the study can be easily translated to other ICPs. The ICP is measured as a relative pressure to the atmospheric pressure and thus the pressure at the outlet of the tube was set to zero.

Diffusion coefficient: The free diffusion coefficient is dependent on the molecule. It has been shown that parameters such as the size, shape and the flexibility have an impact and thus it is difficult to predict the actual diffusion coefficient without measuring it. Furthermore, the diffusion coefficient in the ECS is anisotropic and depends on the location in the brain. For this study a range of different diffusion coefficient in ECS were tested but they are assumed to be isotropic and constant over all the simulated brain tissue. The range assessed was between 10' ¹¹ to 10' ⁹ m ² /s, which includes smaller molecules such as ions and larger molecules like proteins. For simplicity the free diffusion coefficient was used and calculated using an averaged tortuosity = D/D* = 2.

Drug concentration: This is the concentration of drug injected in the tube. The influence of this parameter on the diffusion process was investigated in parts of the study. However, for simulations in which this parameter was not varied, it was set to lOmM.

Elimination rate: The elimination rate is also molecule dependent. In this study a number of elimination rates were tested from 10' ⁴ to 10' ² s' ¹ . However, for simulations in which this parameter was not varied, it was set to 0 in order to remove the influence of this parameter and allow better understanding of the transport mechanisms in operation. Simulation parameters: The boundaries of the brain were set to the ICP and to have a no flux condition. The walls of the tube are set to a no slip and no flux condition. The outlet of the tube was set to atmospheric pressure, which is 0 in relative pressures. The mesh is coarse but with a user-defined maximum element size of 2pm on the curvature (the “U”) of the tube.

Results

Concentration profile and time-dependency

Figure 2 shows the results of a simulation with a 3nL.min flow rate. While some of the drug diffuses through the hole into the ECS, most of it continues along within the tube.

Figure 2b shows the concentration profile along a straight horizontal line from the tube in the ECS. The profile shows an exponential increase with time. The increase in concentration tapers off with time.

Figure 2c shows the amount of drug delivered into the brain in the diffusion-driven mode for a flow rate of 3nL/min.

Effect of flow rate on the transport mechanism

The pressure applied on the brain can be influenced by many different parameters such as the geometry of the tube and the viscosity of the liquid injected. For example, decreasing the tube width while keeping the same flow rate will increase the applied pressure.

However, for the purposes of the current simulations, the size of the tube is fixed and only the flow rate set at the inlet of the tube is changed. The flow rate is optimized to balance between too much pressure, resulting in convection-driven transport and too low pressure inducing backflow.

Figures 3a - 3e show how, respectively, a) the concentration profile, b) the pressure profile, c) the velocity magnitude profile, d) the diffusive and convective flux magnitude, and e) the amount of drug infused into the brain, vary for different flow rates. Figures 3a - 3d show the variation along the same horizontal line as in Figure 2b after 60 seconds; Figure 3e shows the variation with time.

Figure 3a shows that the concentration profiles change with changing flow rates. Figure 3a shows that drug delivery with this device can work in both diffusion and convection mode depending on the pressure regime/flow rate. Figure 3a shows a transition from concave to convex concentration profile (or exponential to inverse s-shape) from l.OnL/min to 7.0nL/min denoting a change in regime from diffusion-driven to pressure-driven. At 3 nL/min the profile has an exponential shape, but for 5 nL/min and 7 nL/min the profile becomes like an inverse S-shaped curve.

These different concentration profiles can be explained by looking at the pressure variation in Figure 3b. At 3 nL/min the pressure in the brain stays almost exactly constant and is only a few Pascal higher than the initial pressure in the brain (dashed line). For higher flow rates, the pressure is increase by a few hundred Pascal near the hole, effectively doubling the ICP just by injecting 7 nL/min into the tube.

Similar profiles can be seen when looking at the velocity magnitude in Figure 3c. Indeed a pressure gradient induces a velocity and moves the drug towards the lower pressure, which is the basis for the convection transport mechanism.

This is confirmed by Figure 3d where the convective flux magnitude is compared with the diffusive flux magnitude. It can be seen that for higher flow rates, such as 5 nL/min or 7 nL/min, the convective flux is higher than the diffusive flux which makes sense as there is a higher pressure gradient, as indicated in Figure 3b. This, in turn, explains the difference in the concentration profiles as it can be proven mathematically that a convection-driven transport is characterised by an inverse S-shaped concentration profile, while a diffusion-driven transport has an exponential-like profile.

When the flow rate is decreased even further, as exemplified by the 1 nL/min line, the pressure inside the tube gets smaller than the pressure in the brain, which creates a backflow and only a very small amount of drug is able to diffuse against this flow and enter the brain. Figure 3e shows the total amount of drug delivered to the brain over time. The convection- driven regime results in a higher overall amount of drugs delivered than a diffusion-driven regime. However, the diffusion-driven regime allows local delivery of small amounts of drugs only to a specific area of need and so results in a much more controlled and focused treatment.

Impact of drug-dependent parameters on the amount of drug delivered over time

To understand the influence of different parameters on the delivery process, these parameters were adjusted and the amount of drug infused in the brain over time recorded. The results are shown in Figures 4a - 4c which show, respectively, the effects of changing a) the diffusion coefficient, b) the injected concentration and c) the elimination rate. Figure 4d is a contour plot showing the amount of drug infused depending on the diffusion coefficient and the elimination rate for a drug concentration of lOmM.

The diffusion coefficient (D*) was varied from 1 x 10' ¹¹ to 1 x 10' ⁹ . Figure 4a shows that the amount of drug delivered increased with the diffusion coefficient. From equation 2 above it can be seen that the diffusion term is directly proportional to the diffusion coefficient. However, multiplying if by 10 from 1 x 10' ¹¹ m ² /s to 1 x IO' ¹⁰ m ² /s makes a much smaller difference in the amount of drug infused than increasing D* from 1 x IO' ¹⁰ m ² /s to 1 x 10' ⁹ m ² /s. However, the shape of the curve is the same for all choices of D*.

The effects of changing the drug concentration Cd are shown in Figure 4b. It can be seen that the total amount of drug in the brain is linearly proportional to the concentration of drug in the tube.

The final drug-dependent parameter considered was the elimination rate. Figure 4c shows the effect of variation in this parameter. The two higher rates in this graph show that the amount of drug infused does not increase linearly, but converges to a steady state quantity. This steady state is reached when the same amount of drug is removed by the uptake, loss or clearance (represented by the elimination rate) as is delivered from the tube into the brain. Whilst a small elimination rate k _e nm = 1 x 10' ⁴ s' ¹ ) makes almost no difference, further order of magnitude increases in the elimination rate significant reduce the concentration of the drug in the brain. Figure 4d is a contour plot showing the amount of drug infused after 60 seconds of simulation for varying combined values of the diffusion coefficient (vertical axis) and elimination rate (horizontal axis). Unsurprisingly, based on Figures 4a and 4c, the highest drug concentration is achieved for high diffusion coefficients and small elimination rates, whilst the lowest drug concentration is achieved by the opposite combination.

Impact of the size and number of holes on the amount of drug delivered over time

Figures 5a and 5b show, respectively, the amount of drug infused in the brain as a function of time with varying a) hole sizes and b) different numbers of holes. Figures 5c and 5d show the spatial distribution of drug from a tube with 3 holes after c) 1 second and d) 60 seconds.

To investigate the effect of these factors, simulations with holes of 5pm, 10pm and 15pm in width (the simulation being in 2D) were computed and the amount of drug infused into the brain plotted as shown in Figure 5a. It is noted that the amount of drug is not linearly proportional to the width of the hole as doubling the hole size does not result in twice as much drug infusing into the brain. Furthermore, increasing the hole size from 5 pm to 10pm has a bigger impact on drug infusion than increasing it from 10pm to 15pm. Similar behaviour is observed with variations in the number of holes, as shown in Figure 5b.

These results arise because the drug concentration gradient is shared by the different holes. As the driving transport mechanism in a diffusion-driven delivery is the concentration gradient, this results in a slower delivery. For example, in the case of 3 holes, Figures 5c and 5d show that even though three different circles can be observed at the very start of the process (Figure 5c, after a short amount of time (and at least by 60 seconds, as shown in Figure 5d, the three circles fuse together forming one big circular distribution and thus only one large-scale concentration gradient around all of the holes rather than three separate gradients.

Summary of simulation results

When all the elimination processes are suppressed, the amount of drug delivered increases linearly once the process has reached a steady-state (i.e. after approximately 1 minute). This is because the source of delivery, namely the concentration in the tube, stays almost constant over time in the whole tube. The concentration profile however, does not increase linearly with time. This can be explained by the drug filling up a 2D circular surface, rather than just a line. Thus, with time, the radius of the concentration of the drug distribution increases proportionally to the square root of the total amount of drug, making the increase reduce as time progresses.

The transport mechanism driving the drug infusion can be either dominated by diffusion or convection. At lower flow rates, the pressure applied to the brain is decreased and, as convection is induced by a pressure gradient, its contribution to the transport mechanism is reduced. Thus, below a certain flow rate, a diffusion-driven process can be achieved. Figure 4d shows that, using the modelled parameters, a flow rate of around 3 nL/min results in a diffusive flux magnitude which is greater than the convective flux magnitude.

This diffusion dominated process can also be observed in the concentration profiles. The diffusion-driven delivery has an exponential-like shape while the convection-driven delivery has an inverse S-shaped curve (i.e. it has a higher concentration extending further from the delivery site).

Conversely, decreasing the flow rate too far will result in a higher pressure inside the brain than in the tube, thus causing backflow. There is therefore a range of optimal flow rates which will produce diffusion dominated delivery. However, the exact range will depend, amongst other factors, on the pressure in the brain, the geometry of the tube and/or the drug delivered.

The effect of changing the diffusion coefficient models the real-world effect of delivering different-sized drug molecules. Bigger molecules, such as proteins, will have a smaller diffusion coefficient and thus a slower delivery, whilst smaller molecules, for example individual ions such as sodium or potassium, have large diffusion coefficients and therefore result in a faster delivery process.

The almost linear relationship between the diffusion coefficient and the total amount of drug delivered arises because of the choice of the diffusion-driven regime and would likely be different at higher flow rates. Changes in the drug concentration were shown to have linear effects on the amount of drug delivered and this is true for across a wide range of drug concentrations. However, taking account of an elimination rate (for whatever biological reason such as uptake, loss or clearance processes) has a significant impact on the shape of the curve with the amount of drug delivered converging to an equilibrium in which as much drug is diffusing out of the tube as is being removed from the ECS. Higher elimination rates result in faster convergence on a lower value.

Changing the number of holes or increasing the size of the holes was found not to have a linear effect on drug diffusion. This is particularly true of increased numbers of holes and can be explained by the interaction of the diffusion process between the holes after an initial period which leads to one large distribution with a lower concentration gradient, rather than separate circular distributions around each hole. If the holes were spaced significantly further apart, then such interactions might be avoided and a linear increase in delivery might result. This interaction may be useful in allowing a bigger and more uniform distribution to be achieved whilst retaining diffusion-driven delivery.

Overall the simulations demonstrate that a diffusion-driven transport regime can be achieved in the brain with a reduced applied pressure and a small delivery tube. In the geometry simulated, specifically a 10pm width tube with a U-shaped geometry, most of the drug continues along the tube and is released to the outlet outside of the brain. This make it possible to avoid almost any over-pressure on the brain. Adapting the flow rate from those starting parameters allows diffusion-driven delivery to be achieved, but the precise flow rates to achieve this will vary, particularly if the size of the tube, the drug or the animal (or region of the animal) being treated differs.

Devices and Systems

Figures 6a and 6b show schematic illustrations of a device 10 according to an embodiment of the present invention implanted in the brain. Figure 6a shows the components of the device at the implanted or distal end, with the structure of the device omitted for clarity. Figure 6b illustrates the entire device, with the distal end magnified so that features can be seen.

Figures 7 and 8 show schematic illustrations of systems 1 according to embodiments of the present invention which incorporate devices 10 such as those in Figures 6a and 6b. The device has a cylindrical delivery tube 20 which forms a U-shape at the distal end of the device with an inlet arm 21 and an outlet arm 22 connecting the U-shaped end 23 to the proximal end of the device and to a supply of drug solution and a waste reservoir respectively (not shown).

In order to allow for implantation through a sub-cm incision, the delivery tube has an outer diameter of around 10-50 pm and, in particular embodiments, an inner diameter of 10 pm.

Near the U-shaped end 23, a plurality of holes 25 are formed in the delivery tube 20. In the examples illustrated in both Figures 6a and 6b, these are provided in a plurality of groups each having a plurality of holes, but it will be appreciated that the configuration of the holes and their spacing from each other can be varied, for example as set out in relation to the simulation described above.

In certain embodiments a semi-permeable membrane is formed over or in the hole(s) 25. This membrane can help to permit diffusive passage of smaller molecules, such as small-molecule drugs, from the interior of the delivery tube 20 to the exterior, whilst preventing backflow of larger molecules, such as proteins, from the region of the body where the device is implanted into the interior of the delivery tube. This can be useful as it avoids the need for over-pressure in the delivery tube to prevent backflow and can ensure that the predominant delivery mechanism for the drug is diffusive rather than convective.

Proximal the U-shaped end 23 and both arms 21, 22 of the delivery tube 20, a plurality of sensing electrodes 30 are arranged. These electrodes can be arranged to sense a variety of localized variables, such as drug concentration, brain activity. All of the sensors may be identical in their nature, or different sensors (or identical sensors with different) purposes may be provided. In some embodiments, sensors can also be provided in the delivery tube 20 itself.

The sensors 30 are connected to the proximal end of the device by conductors 32 which terminate in electrical connections 34. The conductors 32 carry the outputs from the sensors to one or more processors which are located at the proximal end of the device, or may be located remotely from the device with communication modules transferring the outputs from the sensors, either raw or processed, to the processors. If a plurality of sensors of the same type are provided, then these may be arranged to detect the effect of or amount of diffusion of drug from the delivery tube 20.

At the proximal end of the device a pump 40 is connected to pump drug solution from a supply reservoir 41 into the delivery tube 20. The pump may be any known pump suitable for this purpose, such as an electro-osmotic pump or a peristaltic pump.

The device can be arranged so that the pressure applied at the connection to the inlet arm is balanced to the pressure at the connection to the outlet arm. An outlet flow control unit 42 connected to the outlet arm can be used to provide this functionality. The outlet flow control unit may include one or more valves and may also include a further pump. The processor 43 may control the valve(s) and/or pump in order to balance the pressure based on signals received from one or more sensors.

In certain embodiments (for example, as shown in Figure 7), the outlet flow control unit 42 can also separate the contents of the outlet flow into one or more constituent components, potentially alongside a waste output. In other embodiments (for example, as shown in Figure 8), the outlet flow control unit 42 does not separate the outlet flow and it is passed entirely to waste.

This can help to ensure that the pressure in the delivery tube 20 is maintained constant and that the delivery of the drug through the holes 25 is diffusive (or predominantly diffusive) rather than convective.

In some embodiments, such as those shown in Figures 7 and 8, more than one delivery reservoir 41 is provided and connected to the inlet arm 21 of the delivery tube 20. This can increase the flexibility and usefulness of the device. For example, in certain embodiments, one delivery reservoir may contain a drug in concentrated form whilst a second delivery reservoir contains a transport solution. By separately controlling the flow from each reservoir (for example using valves or additional pumps), the concentration of the drug solution being supplied into the inlet arm 21 of the delivery tube can be controlled. In other embodiments multiple delivery reservoirs may be provided each containing a different drug in order to allow for different treatments to be administered through a single device. A further reservoir may contain an inert solution which can be used to flush out any remaining drug from the delivery tube in between administrations.

Manufacture of devices

Devices according to embodiments of the present invention may be manufactured by a variety of methods. Particular methods for manufacturing devices according to embodiments of the present invention are monolithic fabrication or mold fabrication.

Two examples of monolithic fabrication methods are described below with reference to Figures 9 and 10.

In the first example, the steps are shown in Figure 9 and are as follows:

1. A carrier wafer (100) is coated with suitable photoresist or mold material (110).

2. The coated wafer is patterned to make trenches that correspond to the regions for the formation of the monolithic tubes.

3. Conformal polymeric deposition/coating of parylene C (120) is then carried out using techniques such as low-pressure chemical vapour deposition. The thickness of the coating is sufficient to achieve a hermetic seal of the trenches on the top side of the wafer.

4. Metal electrodes (130) are defined through a second lithography process. This step may be replaced by another sensor fabrication option depending on the specific sensor applied, which may include fabricating the sensor inside the microfluidic conduit.

5. The sensor/electrodes are passivated via another polymeric coating (140).

6. The sensor contact areas are etched.

7. Inlet and outlet holes are etched in the microfluidic conduit.

8. The device outline is etched.

9. The mold material (100) is removed/dissolved to release device.

The PEDOT:PSS coating on the surface of the recording sensors (not shown) could be coated on the contact via electropolymerization or some other technique.

The fabrication process does not have to follow the stated order. In the second example, the microfluidic conduits do not have a collocated sensor and the steps are shown in Figure 10 and are as follows:

1. A carrier wafer (100) is coated with suitable photoresist or mold material (110).

2. The coated wafer is patterned to make trenches that correspond to the region for the formation of the monolithic tube.

3. Conformal polymeric deposition/coating of parylene C (120) is then carried out using techniques such as low-pressure chemical vapour deposition. The thickness of the coating is such as sufficient to achieve a hermetic seal of the trench on the top side of the wafer.

4. Inlet and outlet holes are etched in the microfluidic conduit.

5. The device outline is etched.

6. The mold material is removed/dissolved to release device.

Control

Figure 11 shows a flow chart for a control system algorithm according to an embodiment of the present invention. The initial steps in the algorithm (SI -S3) are the initialization of the device and controller. The controller can be implemented by a hardware processor 43 as illustrated in the embodiments shown in Figures 7 and 8. The controller may be located proximate to the proximal end of the device 20, or may be located remotely and in communication with the sensors 30, pump 40 and other components of the system and device in order to receive information and control the operation of the system and/or device. The remote connection of a controller may be by wired or wireless connection of any known type.

The combination of sensors 30 and the ability to delivery drug(s) through the delivery tube 20 and holes 25 allows a processor in communication with, and receiving information from, the sensors 30 to control the delivery of a drug through the delivery tube.

In some embodiments, the processor can process the data from the sensors to determine a physiological condition of the portion of the body where the device is implanted (S4). For example, the sensors may record electrophysiology data which is processed by the processor. This data may be compared to one or more baseline data sets to determine if an abnormality exists (S5). Alternatively, the data may be compared to previously-obtained data to determine and identify a transition or change in the data, which may, for example, be indicative of a transition from a normal state to an abnormal state or vice-versa.

In one embodiment, the device is implanted in the brain of a patient. In such embodiments the processor may be arranged to analyse the electrophysiology data from the sensors to identify and distinguish epileptic seizures or pre-seizure conditions from normal brain activity. Alternatively or additionally the processor may be arranged to identify patterns in the sensor data indicative of transition from a normal state to a seizure state.

The processor can be arranged to control the delivery of a drug through the delivery tube 20 of the device 10 based on the sensor data and the analysis of that data (S6). The processor can also be arranged to control which drug is delivered and the concentration of the drug delivered where the configuration of the device provides for such possibilities (S6). This can include, for example, control of valves between the various reservoirs and the inlet arm 21 of the delivery tube 20 and/or control of different pumps which cause different fluids to flow into the inlet arm 21 and can control the ratio of the fluids being supplied.

For example the processor may be arranged to deliver a drug when an abnormal condition is identified in order to treat the condition. Alternatively or additionally the processor may be arranged to deliver a drug when a transition from a normal state is identified, which may allow preventative treatment to avoid the transition to the abnormal state, or to delay or ameliorate the effects of the transition, or to lower the severity of the abnormal state.

The processor may control to the delivery of the drug by controlling the pump connected to the delivery tube 20. This allows the processor to control the on/off state of a drug solution being passed through the delivery tube as well as the flow rate of that drug solution.

The processor may also use the data from the sensors to identify when the patient returns to a normal condition (S9), or when the patient’s condition stabilises and may be arranged to adjust the delivery of the drug accordingly (S10).

In some embodiments, the sensors provide information to the processor about the delivery of the drug (S6) and this information may be used by the processor to control the delivery (S8) of the drug, for example to reduce or stop the delivery once a certain delivery level (e.g. concentration) has been reached. Alternatively or additionally, the processor may use this information to determine whether the delivery of the drug is diffusion-driven (for example by considering the delivery profile of the drug using a plurality of sensors) and adjust the pump(s) according (S7-S8)

In one particular embodiment the device is arranged to be implanted in the brain and the sensors and processor are arranged to identify and distinguish epileptic seizures or pre-seizure conditions from normal brain activity. On identification of a seizure or pre-seizure condition, the processor controls the pump to cause delivery of an anti-epileptic drug to control the seizure. On detection of the return of a normal brain condition the processor controls the pump to stop delivery of the drug.

Materials

Devices according to embodiments of the present invention can be used for delivery of drugs to many different parts of the body. Although specific examples have been given above in relation to the use of devices to delivery drugs to a human brain, the present invention is not limited to such uses. Further examples include the use for delivery of drugs to malignant tumours - e,g. pancreatic tumor. Others include nerve injury, local pain or inflammation.

The materials of the devices, and particularly the distal end of the devices, are selected for good compatibility with tissue in the region where the device is to be implanted. Apart from the use of generally biocompatible materials such as PDMS, polyimide, parylene, other important considerations for tissue compatibility are geometry and bending stiffness.

In relation to bending stiffness, the distal end of the device should ideally approach the bending stiffness of the target tissue. This bending stiffness can vary considerably for human tissue from around 10' ¹³ - IO' ¹⁰ Nm/mm for the brain, through around 10' ² Nm/mm for muscle, to 1,000s Nm/mm for bones.

Bending stiffness is a combination of geometry (principally thickness) and the Young’s modulus of a material. The forgoing description is exemplary in nature only, and the skilled person will understand that changes and variations on the disclosed embodiments are possible within the scope of the claims. The claims define the invention.

References

[1] D. J. Wolak, R. G. Thorne, Molecular pharmaceutics, pp. 10.5: 1492-1504 (2013)

[2] J. Zhou et al, Proceedings of the National Academy of sciences, pp. 110.29: 11751-11756 (2013)

[3] A. M. Mehta, A. M. Sonabend, J. N. Bruce, Neurotherapeutics, pp. 14.2: 358-371 (2017)

[4] I. U. Khan, C.A. Serra, N. Anton, T. Vandamme, Journal of controlled release, pp. 172(3), 1065-1074 (2013)

[5] O. Lewis, M. Wooley, D. Johnson, A. Rosser, N. U. Barua, A. E. Bienemann, S. S. Gill, S. Evans, Journal of neuroscience methods, pp. 259, 47-56 (2016)

[6] R. H. Bobo, D. W. Laske, A. Akbasak, P. F. Morrison, R. L. Dedrick, E. H. Oldfield, Proceedings of the National Academy of Sciences, pp. 91(6), 2076-2080 (1994)

[7] C. M. Proctor, A. Slezia, A. Kaszas, A. Ghestem, I. Del Agua, A.-M. Pappa, C. Bernard, A. Williamson and G. G. Malliaras, Science advances, p. eaaul291 (2018)

[8] E. McGlynn, V. Nabaei, E. Ren, G. Galeote-Checa, R. Das, G. Curia, H. Heidari, Advanced Science, p. 8(10), 2002693 (2021)

[9] C. J. Bettinger, M. Ecker, T. D. Y. Kozai, G. G. Malliaras, E. Meng, W. Voit, MRS bulletin, 45(8), 655-668 (2020)

[10] A. Di Stefano, P. Sozio, A. lannitelli, L. S. Cerasa, Expert Opinion on Drug Delivery, pp. 6(4), 389-404 (2009)

[11] A. Lecomte, E. Descamps, C. Bergaud, Journal of neural engineering, pp. 15(3), 031001 (2018)

[12] E. Sykova, C. Nicholson, Physiological reviews, pp. 88.4: 1277-1340. (2008)

[13] S. W. Kuffler, D. D. Potter, Journal of Neurophysiology, pp. 27.2: 290-320 (1964)

[14] W. Zhan, C.-H. Wang, Journal of Controlled Release, pp. 271 : 74-87 (2018)

[15] J M. Czosnyka, J. D. Pickard, Journal ofNeurology, Neurosurgery & Psychiatry, pp. 75.6: 813-821 (2004)

[16] W. Zhan, F. Rodriguez y Baena, D. Dini, Drug delivery, pp. 26.1 : 773-781 (2019)

[17] M. E. Rice, G. A. Gerhardt, P. M. Hierl, G. Nagy, R. N. Adams, Neuroscience, pp. 891- 902 (1985) [18] R. G. Thorne, S. Hrabetova, C. Nicholson, Journal of Neurophysiology, pp. 3471-3481 (2004)

[19] R. Raghavan, M. L. Brady, M. I. Rodriguez-Ponce , A. Hartl ep, C. Pedain, J. H. Sampson, Neurosurgical focus, p. 20.4: E12 (2006)

[20] F. T. Sun, M. J. Morrell, R. E. Wharen. 2008, Neurotherapeutics, pp. 5(1), 68-74 (2008)

[21] B. Rosin, M. Slovik, R. Mitelman, M. Rivlin-Etzion, S. N. Haber, Z. Israel, E. Vaadia, H. Bergman, Neuron, pp. 72(2), 370-384 (2011)

All of the above references are hereby incorporated by reference in their entirety.