Membranes for Medical Devices

Abstract

The invention relates to an implantable chamber comprising a closed shell made of a permeable membrane, said membrane comprising at least one layer of porous biocompatible polymer, with pores controlled and homogeneously distributed over the membrane.

Claims

1. A biocompatible implantable chamber, comprising a closed shell made of a membrane, delimiting an inner space, wherein the membrane comprises pores on at least one part of its surface, and wherein the pores are homogeneously distributed on the at least one part of the membrane, and wherein the diameter of the pores is comprised between 200 nm and 100 μm.

2. The implantable chamber of claim 1, wherein the membrane comprises more than one a layer of biocompatible porous polymer.

3. The implantable chamber of claim 1, wherein the membrane consists in one layer of porous biocompatible polymer.

4. The implantable chamber of claim 1, wherein the membrane comprising the homogeneously distributed pores is made from a polymer selected from the group consisting of polycarbonate (PC), polyurethane (PU), polyethylenimine, polyolefins, polyester, fluoropolymers, and polyamides.

5. The implantable chamber of claim 1, wherein the porous biocompatible polymer is made hydrophilic by surface physical or chemical modification, and covered with at least one hydrophilic polymer.

6. The implantable chamber of claim 1, wherein the layer of porous biocompatible polymer has a pore density of between 10.sup.3 pores/cm.sup.2 and 10.sup.9 pores/cm.sup.2.

7. The implantable chamber of claim 1, wherein the total thickness of the membrane is between 5 μm and 250 μm.

8. The implantable chamber of claim 1, wherein the porous biocompatible membrane is covered a hydrophilic polymer which contains at least one biologically active molecule, in particular covalently bound to a layer on the surface of said membrane.

9. The implantable chamber of claim 1, wherein the pores have a diameter between 5 μm and 100 μm.

10. The implantable chamber of claim 1, which comprises at least one connector comprising a body attached to the membrane, and a conduit connected to the connector so as to be in hydraulic communication with the inside of the pouch, which makes it possible to establish a communication between the exterior and the interior of the shell.

11. A kit comprising a. an implantable chamber according to claim 1, which comprises at least one connector comprising a body attached to the membrane, and a conduit connected to the connector so as to be in hydraulic communication with the inside of the pouch b. a catheter that can be connected to said connector at one end and that can be connected to a source of delivery of a compound of interest at the other end.

12. The kit of claim 11, wherein said catheter presents an injection port, in particular implantable subcutaneously, at the end that can be connected to a source of delivery.

13. The kit of claim 11, which further comprises a pump, a needle, a syringe, or a pen, making it possible to send the compound of interest from the source of delivery to the chamber within the catheter.

14. The kit of claim 11, further comprising sensors and/or captors to measure the level of a given biomarker in the blood and optionally send signals to the external source of the compound of interest.

15. A method for administering a substance to a patient in need thereof, comprising administering said substance to the patient's body through a catheter linked to a chamber according to claim 1 wherein said substance diffuses by the pores of the membranes of the chamber.

16. The method according to claim 15, wherein the chamber is implanted extraperitoneally.

17. The method according to claim 15, wherein said substance is a coagulation factor, in particular Factor VII, Factor VIII or Factor IX, for use thereof for the treatment of hemophilia.

18. The method according to claim 17, wherein the chamber is implanted intraperitoneally or extraperitoneally.

19. The method according to claim 15, wherein said substance is a chemotherapy drug (such as Paclitaxel, Taxol, Cis-Platin, cyclophosphamide, 5-fluorouracile, vincristine, or presnisolone) for use thereof for the treatment of cancer.

20. The method according to claim 19, wherein the chamber is implanted intraperitoneally or extraperitoneally.

21. The method according to claim 15, wherein said substance is an anticancer drug for use thereof in the treatment of a cancer, wherein said anticancer drub is preferably selected from the group consisting of temozolomide, bevacizumab, carboplatin, carmustin, mibefradil, afatinib, tandutinib, and enzastaurin.

22. (canceled)

23. The method according to claim 21, wherein the cancer is a brain cancer, in particular glioblastoma.

24. (canceled)

25. The method according to claim 23, wherein the chamber is implanted intra-parenchimally.

26. The method of claim 15, wherein said substance is insulin.

Description

DESCRIPTION OF THE FIGURES

[0231] FIG. 1: Comparison of pores repartition and homogeneity between membranes of the prior art (A) and membranes according to the invention (B, C). Scanning Electron Microscopy (SEM) of track-etched membranes (A) (The pores are in black and the membrane in light color), Optical microscopy of laser cut membrane with pores at 45 μm, Magnificent ×10 (B) and Laser cut membrane with pores at 45 μm. Magnificent ×20 (C). The pores are in light color and the membrane in black for B and C.

[0232] FIG. 2: Permeability of Polyester membrane samples with different pore sizes (0, μm) and thicknesses (E, μm) to insulin. Diffusion time of 24h at 37° C., each point shows result of one sample, horizontal bar shows mean value with SEM as error bars. ** and *** p<0,01 and 0,001 respectively, One-way ANOVA with Tukey post-hoc test.

[0233] FIG. 3: Permeability of symmetric membranes to Humulin-R® and Insuman Infusat® after 24h. Each point shows a replicate, horizontal bar represents mean value and error bars corresponds to SD.

[0234] FIG. 4: Permeability of membrane samples to 3-5 KDa FITC-Dextran after 4 weeks of implantation in extraperitoneal on diabetic rats compared to non-implanted samples. Mean±SEM. *** p<0,001 vs non implanted CTL, T.Test

[0235] FIG. 5: Masson's Trichrome staining reveals presence of vessels in direct contact with membrane candidates tested. Dotted square show area enlarged on right panel. Arrows indicate vessels in direct contact with membrane. M: membrane.

[0236] FIG. 6: Number of vessels and their mean surface in tissues surrounding the three membrane candidates implanted for 4 weeks in extraperitoneal on rats. Mean±SEM. Quantifications were made separately on peritoneal and muscle side, on three non-overlapping field at magnification 10 of Masson's Trichrome stained tissue. * for p<0,05 one-way ANOVA with Tukey post-hoc test as data for peritoneal and muscle side were analyzed separately.

[0237] FIG. 7: Glycemia follow up (A) and corresponding area under the curve (B) after injection of 2 IU of Human insulin in device made of laser cut membranes ExOlin® assembled with flexible membranes implanted in extraperitoneal or after direct IP or SC injection. Repeated measure ANOVA with Tukey Post-test. * For p<0,05 et ** for p<0,01

[0238] FIG. 8: Insulin level in peripheral blood and portal vein at 5 min and 30 min after injection of 2U of Human insulin in devices assembled with membranes implanted in extraperitoneal on diabetic rats.

[0239] FIG. 9: Macroscopic view of new generation of ExOlin® adevice made with laser-cut membrane of the invention the time of retrieving after testing in diabetic rats (total implantation time: 7 weeks), on peritoneal side (A) and muscle side (B). Representative pictures of n=5.

[0240] FIG. 10: Masson's Trichrome staining of tissues surrounding laser-cut membranes at the time of retrieving the device after testing in diabetic rats (total implantation time: 7 weeks). Pictures of muscle side are presented on the upper panel and of peritoneum side on lower panel. Black arrow highlights vessels close to the membrane, M: place device's membrane. Representative pictures of n=5.

EXAMPLES

Example 1. Characterization of the Membranes' Surface in Comparison to the Prior Art

[0241] Scanning Electron Microscopy (SEM) analysis shows a non-homogeneous distribution of the pores generated when using the track-etching process (as used in WO2015086550, WO2016184872 or WO2018087102). In comparison, the pores obtained using the laser technique are very reproducible and do not overlap thus increasing the repartition of the local constraints leading to a better mechanical properties (FIG. 1).

[0242] It was also noticed that, in the conditions used, the pores obtained using the track-etching technique were smaller than the pores obtained with the laser cutting technique and were thus observed using SEM analysis, whereas the pores obtained with laser were observed using optical microscope.

Example 2. Mechanical Characterization of Membranes According to the Invention

[0243] The thickness of the 4 membranes that were tested varied from 40 μm to 125 μm (40 μm, 50 μm, 85 μm and 125 μm considered) with respectively 45 μm, 37 μm, 55 μm and 37 μm of pores diameters.

[0244] The test was performed using a tensile testing machine (MTS 1/M machine) equipped with a 500N load cell and a pneumatic clamping system adapted to the thinness of the membranes. No extensometer is available for this type of specimen (crosshead displacement taken in account). The strain rate was set to be 0,005s.sup.−1.

[0245] All samples tested showed elongation at break comprised between 15 and 20 mm versus only 5 mm for the membrane of the prior art (obtained by track-etching). This reveals a high elasticity (and therefore flexibility) of the material tested.

[0246] It was noticed that the evolution of the maximum force is relatively linear as a function of the thickness of the membrane, except for the membrane of 85 μm thickness because the structure of this membrane is different (larger pores) than the others. When comparing the membranes having 85 μm and 125 μm thicknesses (but different pore sizes), it was noted that the results were very similar considering either force or elongation (stress or strain).

[0247] From 85 μm thickness, the membranes are stronger that those made of PET with track-etching. Indeed, the membranes are 4 times stronger than those from the prior art. Moreover, they can be more distorted (about 4 times more), but their stiffness is lower (lower Young's modulus). The membranes with a lower thickness (40 μm and 50 μm) underwent a lower force than those made of PET with a thickness 100 μm, but they can be more distorted.

[0248] In view of these results, it was decided to further investigate the two membrane thicknesses (85 and 125 μm) for the development of a device suitable for mini-invasive surgery.

[0249] It was also noticed that mechanical resistance at break of the 1-layer laser cut membrane with homogenously distributed is at least as good as the 3-layers track-etched membrane from previous art.

Example 3 Diffusion Tests of Molecules of Interest Through Membranes

[0250] Discs of membranes were cut at a diameter of 22 mm to fit the inner dimensions of a diffusion chamber developed to perform such testing. The diffusion chamber is composed of two compartments which can be screwed together using a PTFE O-ring to ensure the tightness of the chamber once assembled. The lower compartment is filled with a saline solution at 9 g/L when the insulin diffusion is tested. Once the lower compartment is full, the 22 mm disc of membrane is put in place and maintained by placing PTFE O-ring above. The upper compartment is then screwed and filled with the insulin solution to be tested. A lid is placed avoiding excessive evaporation.

[0251] The diffusion chamber is placed at 37° C. during 24 hrs. Afterwards, a sample of medium in upper compartment is taken and the remaining is discarded. The upper compartment is then unscrewed and the disc of membrane is removed. A sample is collected also in the lower compartment and is assessed using BCA assay together with sample from upper compartment.

[0252] Then, permeability is calculated as the ratio of insulin found in lower compartment on total quantity of insulin in both upper and lower compartment. Since the testing is perform in static conditions, concentrations of upper and lower compartment tend to equilibrate if membrane is permeable to the molecule. Therefore, it is important to note that the maximal permeability that can be reached is 50%.

[0253] As first test with membranes displaying a thickness of 125 μm and a pore diameter of 37 μm showed high permeability to insulin. Three other membranes were tested. They displayed different pore diameter (Ø, μm) and thickness (E, μm) (FIG. 2).

[0254] Results obtained after 24h diffusion test with insulin showed very high permeability of all the candidates. The highest permeability was obtained for membranes Ø45/E40 (51.13%), which is significantly higher than the three other candidates. This clearly highlights the decrease of insulin permeability with thicker samples. This trend is confirmed by the statistically higher permeability of membrane Ø37/T50 compared to the sample with same pore size but higher thickness of 125 μm (49.06±0.34% vs. 47.00±0.62%) (FIG. 3). Despite significant differences between the four samples tested, permeability values are all higher than 40% and superior to those obtained with membranes from the prior art.

[0255] Results show that permeability of symmetric membranes to injectable insulin Insuman Infusat® (Sanofi) is comparable to permeability obtained with concentrated insulin Humulin-R® (Eli Lilly), used routinely for quality controls performed at reception of membranes. Permeability after 24h are respectively 34.4±2.0% and 35.3±2.1% for Humulin-R® and Insuman Infusat® (p=0.3923 Unpaired T.Test).

Example 4: Testing of Membranes' Biointegration in Rats

[0256] For each rat to implant, two discs of the same membrane (Ø45/E40, Ø55/E85 or Ø37/E125) were cut at a diameter of 22 mm using a specific cutter. They were then placed in sealed pouch and sterilized with ethylene oxide.

[0257] Two discs were then implanted in extra-peritoneal site on healthy male Wistar rats (n=6 per membrane). One disc placed on the left side of the abdomen (for diffusion tests with 3-5 kDa FITC Dextran for 24h) and one on the right side (for histological and SEM analysis), to ensure having enough material to perform tests after retrieving. Implantation time was for 1 month.

[0258] Membrane discs implantation was performed following these steps: [0259] Make a 2 cm vertical incision using scalpel on the midline of abdomen [0260] Using fine scissors, separate skin from muscles [0261] On the side of linea alba, gently vertically incise the muscle with scalpel until peritoneum is exposed. Make sure incision is large enough to fit with membrane discs which have a 22 mm diameter. [0262] Dissect an extraperitoneal pouch using curved fine scissors and/or forceps. Be careful during dissection to avoid wound in the peritoneum. [0263] Once pouch is large enough, inject sterile saline solution and insert membrane disc in the pouch. Make sure disc is not folded on the site of implantation. [0264] Suture muscle using absorbable 4-0 thread (Vicryl Rapide-Ethicon), by sinusoid movement [0265] Suture skin using Donati pattern using absorbable 4-0 suture thread.

[0266] Animals implanted with membrane discs were kept for 4 weeks prior sacrifice by intraperitoneal injection of Pentobarbital (Euthasol® VET, 182 mg/mL) to retrieve membranes and surrounding tissues.

[0267] One of the two membrane disc was separated from surrounding tissues then rinsed in saline solution to perform diffusion test with 3-5 KDa FITC Dextran (size comparable to insulin) and determine if implantation period affected permeability of the sample. A small membrane samples was cut on the other disc for SEM analysis and the rest was taken with its surrounding tissues for histological analyses. Tissues with membranes were rinsed in saline solution then fixed in buffered formalin for 72h.

[0268] The three selected membrane candidates were implanted for 4 weeks in EP on non-diabetic rats, to investigate their biointegration, and permeability after implantation.

[0269] Membranes were macroscopically observable through peritoneum, at the time of sacrifice, revealing absence of strong fibrotic reactions. Membranes were not folded but difficult to separate from surrounding tissues, requiring to pull hard with forceps. Once separated form tissues, organic deposit stayed on the membrane and appeared to be pass through the pores.

[0270] These macroscopic observations were confirmed by SEM analysis that showed homogenous covering of membranes by organic deposit. The covering was identical for the three samples tested and specific surface patterns of membranes were hardly visible.

[0271] In addition to surface observation, permeability of membranes after 4 weeks of implantation in EP on rats was assessed with 3-5KDa FITC-Dextran. Choice of this molecule rather than insulin is explained by the presence of strong organic deposits that would interfere with BCA Assay validated for insulin titration.

[0272] In accordance with macroscopic observations and SEM analysis, results (FIG. 4) follow the same trend for the three membrane candidates. A significant decrease (p>0,001 for all candidates) of permeability to 3-5 KDa FITC-Dextran is observed after 4 weeks of implantation compared to non-implanted material. Overall, permeability value on implanted membranes is half of the value obtained with non-implanted control (1.72-; 2.00- and 2.13-fold decrease for Ø45/E40; Ø37/E50; Ø55/E85 and Ø37/E125 respectively).

[0273] One of the two discs implanted for 4 weeks in EP on rats was not separated from its surrounding tissue and the whole sample was embedded in paraffin for histological analyses.

[0274] Regardless the membrane candidate implanted, they do not affect global tissue's organization both on muscular and peritoneal side, as shown by Hematoxylin-Eosin staining. Low cell infiltration was also observed together with very low fibrosis, indicating good acceptance of membrane in extraperitoneal. It is worth to mention that increased fibrotic tissue is observed on the edge of membrane discs, as observed with Masson's Trichrome especially for Ø45/E40 and Ø55/E85 candidates (right panel, upper and middle picture). In accordance with organic deposit found on membrane separated from tissues at retrieving, we observe a total enclosing of membrane in tissues. Tissue also passed through the pores, which explains the difficulties encountered to separate membranes and tissues.

[0275] Vascularization around membranes is a critical point. Therefore, this parameter was deeply analyzed.

[0276] FIG. 5 shows pictures of vascularization in the area where membranes were implanted. In enlarged pictures from the left panel, one can observe small vessels that are very close to the membrane material (circular grey structures that seem to detach from the slide) as highlighted by black arrows. For Ø55/E85 and Ø37/E125 candidates, small vessels are observed directly attached on the membrane fibers and even present in some pores. For Ø45/E40 membrane, the vessels appeared to be more distant from the membranes and seem to be mainly located in fibrous tissue (stained in blue) close to the membrane material. These data could mean that the thickness of the membrane could have an impact on healing process as the distance between membrane and vessels is more important with the smallest thickness.

[0277] To have a more precise idea of vascularization tissue surrounding membrane candidates after 4 weeks of EP implantations, quantifications of vessel number and area was performed on slides stained with Masson's Trichrome staining.

[0278] Density of vessels is comparable on both peritoneal and muscle sides for membrane candidates Ø55/E85 and Ø37/E125. For Ø45/E40 membranes, density tends to be higher compared to the other candidates on peritoneal side (14.00±1.41 vs. 10.07±2.98 and 11.17±1.36 for Ø55/E85 and Ø37/E125 respectively). This trend is stronger on muscle side and vessels number is statistically higher with Ø45/E40 compared to Ø37/E125 candidate (p=0.0247, 20.44±4.75 compared to 9.11±2.32) (FIG. 6).

[0279] Similar to vessels' density, mean vessel area is similar on both sides of the membrane, between Ø55/E85 and Ø37/E125. These mean values are also comparable to the one obtained with Ø45/E40 on muscle side. However, for this candidate, vessel area clearly tends to be lower than with Ø55/E85 and Ø37/E125 (220.2±24.7 for Ø45/E40 vs. 403.8±84.57 for Ø55/E85; p=0.0565 and 369.1±31.7 for Ø37/E125; p=0.1312). This reveals smaller vessels on peritoneal side for Ø45/E40 candidate compared to the two others.

[0280] Taken together, these data on vascularization show a higher number of vessels in tissues surrounding Ø45/E40 membrane with a smaller size. This points out that vascularization is less mature with this candidate compared to the others.

Example 5: Efficacy Study in Diabetic Rats of Devices Made from Membranes

[0281] Since diffusion properties performed after the first implantation in rats were impacted with tissues deposition. Consequently, further implantations were realized with the whole device assembled with the same membranes on diabetic rats (induced by streptozotocin) to assess the effect of insulin despites tissue deposition. Diabetic rats under insulin therapy were implanted in EP (extraperitoneally) with device made of laser cut membranes with homogenously distributed pores (37 μm pores and thickness of 125 μm). After a device pre-implantation period of 6 weeks, insulin therapy was stopped and injections of 21U of insulin (Insuman Infusat®*, 21U) were performed. Each rat was successively injected with insulin in device made of laser cut membranes implanted in EP, in SC and in IP, with 48h washout period between each injection. After injection a 2h glycaemia follow up was performed with blood samples at 0, 30, 60 and 120 min to measure Human plasma insulin levels. Following these three injections, rats received a last injection of 2 IU of Human insulin in the device in EP and blood samples were taken at 5 and 30 min post injection in both portal and tail vein.

[0282] Finally, animals with the device made with laser-cut membranes were sacrificed to macroscopically observe its biointegration and integrity, and histological analyses were performed on tissues surrounding the device.

[0283] FIG. 7A shows a faster glycemia decrease after injection of insulin in devices assembled with MEMBRANES, compared to direct injection of the same quantity of insulin in SC or IP. This results in a significantly lower AUC after injection through the device in EP compared to the two other conditions (p=0.0009 vs. SC injection; p=0.0204 vs. IP injection).

[0284] Results of FIG. 8 clearly demonstrate that use of the device made of laser cut membranes in EP on diabetic rat model results in a rapid passage of insulin to portal vein, with significantly higher levels compared to peripheral blood at both times tested (p=0.0193 at 5 min and p<0.0001 at 30 min). This distribution pattern of injected insulin tends to mimic physiological situation and is comparable to what was observed with device made of membranes from the prior art.

[0285] At the end of the study, corresponding to a total time of 7 weeks of implantation, device did not show any folding and kept its integrity at the implantation site. Device was clean and significantly easier to retrieve compared to membranes alone. No thickening of peritoneum was observed and vessels were visible close to the membranes (FIG. 9A). On the muscle side aspect of tissue was also normal, without visible fibrotic reaction (FIG. 9B).

[0286] Histological analyses confirmed macroscopic observations, with very low fibrosis in contact with device's membrane on both muscle and peritoneal sides (FIG. 10 left panel). In addition, vessels were observed on very close to the membranes and even in direct contact with membrane material (FIG. 10 right panel).

[0287] Taken all together these results indicate that the device made with the laser-cut membranes is as efficient as previous version of the device. In addition, both material and topography of the membranes used provide a very good integration at extraperitoneal site in diabetic rats.

Example 6: Laparoscopic Procedure Testing

[0288] Devices designed with laser cut membranes of 125 μm with pores at 37 μm were tested using laparoscopic tools in pigs. The diffusion chamber is inserted into the intraperitoneal space of the animal using a 12 mm outer diameter trocar. Then the diffusion chamber was placed at the desired implantation site (i.e. submucosa, intra-abdominal). Diffusion chamber was also placed in the preperitoneal space by gently incising the peritoneum and suturing this latter when diffusion chamber was placed (transabdominal preperitoneal (TAPP) procedure).

[0289] In comparison with the diffusion chamber of the prior art, the chamber made with laser cut membranes bring real advantages for the minimally invasive surgery because of its flexibility and stiffness allowing respectively to pass through the 12 mm diameter trocar but also unfold when in the intraperitoneal cavity without the need of additional tools, whereas the membranes previously described were impossible to roll to pass through trocar, without damaging them.

[0290] Moreover, the connection system of the chamber of the prior art was conserved and provides an easy grip for the grasper tool. In combination with the flexibility of the laser cut membranes, there is no need to develop additional apparatus, equipment nor tools to roll, inserted, unroll the diffusion chamber.