Carbon Block/Filtration Bed/Conical Reactor with Fluidized Bed System Allowing Small Sorbent Particles to Regenerate Fluid During Extracorporeal Blood Treatment

Abstract

Methods and devices for powdered sorbent regeneration of biologic fluids are disclosed. The present invention includes three novel methods, which may be used singly or in any combination, for constraining or immobilizing powders so that they can be perfused with a biological fluid or dialysate: a porous carbon block filter, a filtration bed of very fine powder, and a cone-shaped reactor.

Claims

1. A method of regenerating a biologic fluid during extracorporeal blood treatment comprising the steps of: (i) pumping blood from a patient through a dialyzer; (ii) withdrawing the biologic fluid from the top of the dialyzer; (iii) pumping the biologic fluid through a dialyzer circuit, and into fluid regeneration system, wherein the system comprises a solid block reactor (SBR), containing a solid carbon block of active carbon; (iv) pumping the biologic fluid from the reactor to a fluid bag; (v) pumping fluid out of the fluid bag and back to the dialyzer; and (vi) changing or replacing the carbon block, the fluid bag, or both, as needed.

2. The method of claim 1 wherein the fluid regeneration system comprises a filtration bed of sorbent particles; and a conical reactor placed below the solid filtration block to create a fluidized bed of sorbent particles.

3. The method of claim 2, further comprising the step of passing the fluid through an inlet in the fluid regeneration system, so that the fluid fluidizes the bed of sorbent particles, and passes through the solid carbon block before exiting through an outlet of the fluid regeneration system.

4. The method of claim 2, further comprising the step of creating the filtration bed of a very fine powder by passing a fluid containing suspended sorbent particles through the filtering surface of the solid carbon block, and then holding the sorbent particles in fixed position by continued fluid flow.

5. The method of claim 2, further comprising the step of using the filtration bed to position or immobilize powdered sorbent particles on the outside of the carbon block during fluid flow.

6. The method of claim 3, further comprising the step of using the filtration bed to allow particles of a few microns in diameter to be used for perfusion and depuration like a column, and provide even flow distribution within the layer of very small particles.

7. The method of claim 6, wherein the carbon block restrains very small particles or fines of other sorbents besides charcoal.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0145] The above-mentioned and other features of this disclosure, and the manner of attaining them, will become more apparent and the disclosure itself will be better understood by reference to the following description of embodiments of the disclosure taken in conjunction with the accompanying drawings, wherein:

[0146] FIG. 1Typical Active Carbon Pore Structure Schematic

[0147] FIG. 2Effects of Particle Size on Sorption Kinetics

[0148] FIG. 3Extracorporeal Blood Treatment System Using Suspension of Active Carbon

[0149] FIG. 4Theoretical Treatment Efficacy as a Function of Plasma Flow Rate

[0150] FIG. 5Fractal Carbon Spheres

[0151] FIG. 6AExample of a Solid Block Active Carbon Filter

[0152] FIG. 6BExample of a Solid Block Active Carbon Filter

[0153] FIG. 7Solid Carbon Block Flow and Holder/Reactor

[0154] FIG. 8Example Insertion Point of a Solid Block Carbon Reactor to an Existing Disposable Kit for Hemodialysis (B Braun Diapact CRRT Machine) (From Diapact manual)

[0155] FIG. 9SBR Disposable

[0156] FIG. 10Auxiliary Priming Disposable

[0157] FIG. 11AComparison of Biologic DT Circulating Active Carbon Suspension with Solid Block Active Carbon Reactor Using Aqueous Dialysate

[0158] FIG. 11BComparison of Biologic DT Circulating Active Carbon Suspension with Solid Block Active Carbon Reactor Using Aqueous Dialysate

[0159] FIG. 12APerformance Comparisons Between Solid Block Carbon and Other Carbon Forms

[0160] FIG. 12BPerformance Comparisons Between Solid Block Carbon and Other Carbon Forms

[0161] FIG. 13Results of Gamma Irradiation of Carbon Blocks

[0162] FIG. 14Conventional CVVHD

[0163] FIG. 15Modification of Conventional CVVHD Using a Carbon Block

[0164] FIG. 16Combination of Conventional and Carbon Block Methods with Infusate

[0165] FIG. 17Addition of Effluent Pump and Reservoir

[0166] FIG. 18ACalcium Phosphate Powder Without Fluid Flow

[0167] FIG. 18BCalcium Phosphate Powder With Fluid Flow

[0168] FIG. 19ADifferences between a standard column and the carbon block/filtration bed approach

[0169] FIG. 19BSurfactants in the fluid may possibly be included in the fluid to aid in meeting particle size, fluid density and viscosity, other fluid characteristics, fluid/particle affinity, surface tension, etc.

[0170] FIG. 20Diagram of the heating circuit of the ThermalCore HT System which includes the DeBakey roller pump and BioTherm heat exchanger, and the NxStage sorbent-dialysis system.

[0171] FIG. 21CCS Test Apparatus

[0172] FIG. 22Outlet Temperature over Time.

[0173] FIG. 23CBFB Pressure over Time (mmHg)

[0174] FIG. 24CBFB Hydraulic Resistance over TimemmHg/(mL/min)

[0175] FIG. 25Cone Reactor Hydraulic Resistance over Time (mmHg)

[0176] FIG. 26Cone Reactor at Startup

[0177] FIG. 27Cone Reactor Soon after Startup

[0178] FIG. 28Mature Cone Reactor

[0179] FIG. 29Particle Cloud at Highest Point

[0180] FIG. 30Reactor at End of Experiment 1

[0181] FIG. 31Early CBFB Flow Uniformity Test

[0182] FIG. 32CBFB Flow Uniformity Test at End of 4h Run

[0183] FIG. 33CBFB Weight Gain Over Time

[0184] FIG. 34Experiment 2 Total Inlet Pressure over Time

[0185] FIG. 35Experiment 2 Flow

[0186] FIG. 36Cloud In Cylinder Acts As Particle Size Classifier

[0187] FIG. 37CBFB Flow Uniformity

[0188] FIG. 38Experiment 3 Setup

[0189] FIG. 39Experiment 3Simulated Blood Flow through Biotherm/MCH-1000 over Time

[0190] FIG. 40Experiment 3 CBFB Inlet Pressure over Time

[0191] FIG. 41Experiment 3 Blood Pressure over Time

[0192] FIG. 42NxStage TemperatureTrace is Output Temperature of Blood to Reservoir

[0193] FIG. 43Unitary CCS

[0194] FIG. 44 BioLogicHT Circuit Schematic

[0195] FIG. 45The Combined system

[0196] FIG. 46ASpreadsheet, First worksheet (part 1 of 2): The output for Funnel Reactor Design Calculations

[0197] FIG. 46B1Spreadsheet, First worksheet (part 2 of 2; top): The output for Funnel Reactor Design Calculations

[0198] FIG. 46B2Spreadsheet, First worksheet (part 2 of 2; bottom): The output for Funnel Reactor Design Calculations

[0199] FIG. 47Spreadsheet, Second Worksheet: The summary of the Funnel Reactor Design calculations

[0200] Corresponding reference characters indicate corresponding parts throughout the several views. Although the drawings represent embodiments of the present disclosure, the drawings are not necessarily to scale and certain features may be exaggerated in order to better illustrate and explain the present disclosure.

DETAILED DESCRIPTION OF THE EXEMPLARY EMBODIMENTS

[0201] The embodiments disclosed below are not intended to be exhaustive or limit the disclosure to the precise forms disclosed in the following detailed description. Rather, the embodiments are chosen and described so that others skilled in the art may utilize their teachings.

[0202] We envision the combined system to be arranged as in FIG. 45, so that particles which are too small to stay within the fluidized bed travel upwards to form the filtration bed around the carbon block. When fluid flow is stopped, the particles falling off of the carbon block will return to the fluidized bed.

Section A: Carbon Block for Toxin Removal from Biological Fluids

[0203] The use of porous solid block active carbon results in the rapid sorption kinetics necessary for effective and efficient toxin removal, coupled with other desirable features such as easy constraint of the carbon, mechanical simplicity and low cost.

[0204] Laboratory data show that porous active carbon block is generally equal to or superior to alternative sorption systems using active carbon.

[0205] Additionally, the author has a developed novel method of priming a reactor using the above invention which excludes harmful air and permits rapid and easy insertion of such a reactor into an existing treatment system. The method consists of evacuating the reactor to a high vacuum. When the user fills the reactor from a standard IV bag, the reactor is immediately ready to use without otherwise difficult to remove entrained air.

[0206] The invention consists of a novel application of an existing product to the problem at hand. The existing product is the common solid-block carbon water filter cartridge. The novel application is to apply the solid block carbon filter to the field of extracorporeal blood treatments, including, but not limited to: [0207] HemoperfusionDirect adsorption of toxins from blood [0208] Plasma treatmentadsorption of toxins from patient plasma [0209] Dialysate purification and regeneration [0210] Single pass purification of dialysate prior to entering the dialyzer [0211] Recirculating dialysate purificationthe dialysate fluid is purified after acquiring toxins across the dialysis membrane and thence sent back to the dialyzer after the carbon has adsorbed the toxins. [0212] Recirculating dialysate purification where the solid block carbon purifies the dialysate from tap or other water prior to beginning of treatment [0213] Peritoneal dialysate purification and regeneration [0214] Single pass purification of dialysate prior to entering the dialyzer [0215] Recirculating dialysate purificationthe dialysate is purified after acquiring toxins across the peritoneum and thence sent back to the peritoneum after the carbon has adsorbed the toxins. In this application, an added value of the carbon block is that it will filter white cells and fibrin material from the peritoneal fluid, thus keeping the fluid very clear on outflow from the peritoneum. This may diminish the tendency for obstruction of inflow and outflow catheters. [0216] Recirculating dialysate purification where the solid block carbon purifies the dialysate from tap or other water prior to beginning of treatment [0217] Purification of other circulating fluids such as albumin or plasma when used in a dialysis or plasmapheresis circuit or other extracorporeal blood treatment device.

[0218] Multiple vendors produce solid block carbon filters for both commercial and consumer use. See FIGS. 6A and 6B for examples, including the KX-5carbon block from Matrixx, Inc.

[0219] These solid block filters are solid only in the sense that the active carbon is a single piece; they are actually porous with nominal mean pore sizes typically in the 0.5 to 10 m range. They are made by taking pulverized active carbon, mixing it with a binder and extruding or otherwise processing it into a hollow cylinder. Fluid passes through the block, typically from the outer perimeter of the cylinder, through the active carbon matrix and thence to the center hole. Although this flow arrangement is generally satisfactory and results in minimal flow resistance, there is no reason why other geometries cannot be used to produce other combinations of hydraulic characteristics and column adsorption characteristics. One example of an alternate geometry would be a solid cylinder used in a manner similar to a classic packed column where flow is from top to bottom.

[0220] It should be noted that it is the bare carbon block that is of interest here. Other accessory components of a typical cartridge such as end caps, sealing rings, preliminary filter wrapping, etc., may or may not be useful in specific applications and such may be used or omitted as desired.

[0221] A very common use for these filters is for whole-house residential drinking water filtration. They remove sediment and other particulates (such as sand from a well) by virtue of their porous structure. They also adsorb undesirable taste and odor causing substances by virtue of the active carbon which makes up the porous structure. Some are also rated to remove certain toxins such as lead or chlorine. Rated flow can be in the 11 gpm range. Some carbons are also capable of chemisorption as well as the more usual van der Waals sorption.

[0222] Although the sorbent capabilities are of primary interest here, in some applications, the filtration function may be useful as well.

[0223] Such filters typically have the following characteristics: The carbon is a whole piece. Thus, no means is required to constrain the carbon from leaking out from the reactor as fluid passes through it. The geometry is complex, resulting in fine features as would be the case with powdered active carbon. The mean effective pore length is thus small, resulting in rapid sorption kinetics. They are available in different nominal pore sizes. They are relatively inexpensive, often mass-produced consumer items. Application is simple. The reactor to contain the solid carbon block need only admit fluid to the outer perimeter, collect it from the center hole and seal the ends of the block. (See FIG. 7). They are typically designed for high water flow rates. The hydraulic resistance thus presented even to albumin or plasma at normal dialysis flow rates is modest, typically <200 mmHg.

[0224] A Solid Block Reactor (SBR) includes a block of porous, solid block active carbon, along with a suitable container which seals the ends and allows proper fluid flow. The SBR will typically also include other features such as test, evacuation and fill ports, mounting appurtenances, labels, etc.

[0225] FIG. 8 shows an example of how the solid block reactor (SBR) containing a solid carbon block of active carbon could be used with an existing hemodialysis system. From the patient on the left, blood is pumped through a dialyzer and then returned to the patient. The dialyzer circuit withdraws dialysate from the top of the dialyzer, pumps it through the SBR and thence to a fluid bag. A third pump pumps dialysate out of the fluid bag and returns it to the dialyzer. The difference in flow between the two dialysate pumps creates ultrafiltrate (or an infusion).

[0226] It was found in laboratory testing that the surface tension of fluids tends to permit air to be entrained in the porous active carbon block for some considerable amount of time after liquid flow has begun. Such air has at least two undesirable effects. First, in some machine configurations, this air could be returned to the dialyzer or plasma filter, and thus, for some blood filtration devices, to the patient's bloodstream. In severe cases, this could cause air emboli. Secondly, air removes the carbon which it contacts from active participation in toxin sorption.

[0227] The novel solution to the air is to evacuate the reactor containing the active carbon. A vacuum of 25 mm Hg vacuum or better has proven very effective. Prior to use, the reactor is spiked and filled with one or more bags of priming fluid. Standard sealed IV bag spike ports are used in the reactor so the vacuum is maintained as bags of fluid are connected. It is important that aqueous priming fluid be used to prime the carbon in all cases since proteinaceous fluids such as albumin tend to foam in the presence of air. An example of such a disposable kit ready for vacuum priming is shown in FIG. 9.

[0228] An additional benefit is also obtained. Extracorporeal treatment machines are primed before use by filling the entire fluid pathway with fluid prior to connection to the patient. The goals of priming are to exclude air from the circuit, and typically, check out the machine operation and discharge any impurities in the fluid circuit.

[0229] With an evacuated SBR, the SBR is easily filled with fluid from standard IV bags. Thus, the host machine is primed normally, and the now fluid-filled SBR is then inserted prior to treatment without need for changes in the machine's existing operating protocol. The SBR thus becomes simple to install and apply.

[0230] It should be noted that vacuum assisted priming is only an option, not necessarily a requirement; most carbon blocks cease to emit air after 30 to 45 minutes after start of priming at 200 mL/min and would take less time at higher flow rates.

[0231] As shown in FIG. 10, an orifice may be provided to prevent the SBR from exhausting an IV bag faster than the operator can react to close off a nearly empty bag and open a fresh bag. This prevents air in the bag from entering the SBR.

Laboratory Results

[0232] Presented below are comparisons between an active carbon SBR and other methods of utilizing active carbon as they apply to different treatments. These data serve to illustrate the benefits of the SBR and demonstrate that the use of an SBR does not result in any significant decrease in therapeutic performance.

[0233] The Biologic DT has, as a major labeled use, the treatment of patients with acetaminophen overdose. This machine uses a circulating suspension of powdered active carbon on the dialysate side of a flat plate dialyzer. The results are compared with a laboratory test using an SBR. No dialyzer was used in this test; the load on the SBR active carbon was thus the more severe.

[0234] In the data in FIG. 11B, the Theoretical Perfect Block Reservoir shows the concentration of acetaminophen in aqueous dialysate using a simulated 40L patient if the active carbon were a perfect black hole for acetaminophen. The Theoretical Kd=0.85 Reservoir is the same data, but assuming the use of a dialyzer in the circuit in order to make the data comparable to the graph on the left. As may be seen, there is massive improvement over the older circulating suspension technology. (Note graph scales carefully.)

[0235] The SBR was compared with granules and other forms of carbon using highly mesoporous carbon block. The dwell time, the time between when fluid enters the reactor and when it exits, was seven minutes. Dwell time is simply the volume of the SBR divided by the flow rate. Three different markers, Methylene Blue (MW=320), Albumin (MM65 K) and Blue Dextran (MM2 M) were used with aqueous buffer and a fourth, Bilirubin (MW=585) was used with bovine plasma. In that graph, the Nanofiber was similar to KX Industries' Plekx material.

[0236] Finally, the SBR was compared with some cytokines (various MW) which are implicated as sepsis mediators. Note that the HSDG was optimized for cytokine removal.

[0237] The results are shown in FIGS. 12a and 12b. Clearly, the application of porous solid block carbon to extracorporeal blood treatments is beneficial and useful.

[0238] Another clinical application of the carbon block will likely be in a dialysate regenerating circuit that is used to remove toxins which develop during whole body hyperthermia (a potential therapy for cancer). This will be done to provide the same chemical function we had with the BioLogic-HDT system, which used powdered charcoal, a cation exchanger and a precipitated calcium phosphate as a sorbent suspension (described in Section C). We have collaborated with the KX Company to create carbon blocks containing the same powdered carbon as was included in our BioLogic-HDT system (derived from coconut shells). We have tested these carbon blocks and shown that binding of toxins is essentially the same as for the powdered carbon in suspension. The carbon block we plan to use will be packaged in a clear plastic housing, evacuated of air, and be clean but not sterile. It contains about 300 grams carbon and is approximately 10 long and 2.5 in diameter.

Section A-2: Special Applications of Sterile Carbon Block

[0239] As described above, regeneration of dialysate during standard hemodialysis can be performed using sorbents which are clean but not sterile. In a treatment of 3-8 hours, bacterial proliferation is not a problem. However one type of hemodialysis therapy is performed for very long periods, up to 72 hours, and therefore utilizes sterile dialysate. This type of dialysis is called Continuous veno-venous hemodialysis or CVVHD. A variation of this therapy is Continuous veno-venous hemofiltration or CVVH. In this therapy, the removal of toxins is by hemofiltration (convection) across the dialyzer membranes, and sterile fluid is infused to the blood to replace the filtered fluid. Here also, sterile fluid is used for infusion to the blood returning to the patient. A combination of hemodialysis and hemofiltration techniques is also used, called CVVHDF. In any of these applications, if the dialysate or hemofiltrate is to be regenerated by carbon block, the carbon block must be sterile and provided within in a sterile perfusion cartridge.

[0240] Gamma radiation is a practical method for sterilizing many medical devices. Carbon is relatively insensitive to gamma radiation; it is used extensively in nuclear applications as a neutron moderator in nuclear reactors and in high-energy particle accelerator installations to receive beam dumps. Carbon blocks that are made with binder typically use either polypropylene or polyethylene in various molecular weight formulations. The latter has been shown to withstand gamma radiation doses up to 1000 kGy. It is thus a reasonable expectation that carbon blocks may be successfully sterilized by gamma radiation, but there does exist some concern that the pore structure might be modified by the sterilizing gamma dose. To test this concern, four carbon blocks were tested, two of which received a gamma dose measured to have been between 35.62 and 37.84 kGy over 1000 minutes.

[0241] As may be seen in FIG. 13, there is not a significant difference in creatinine adsorption performance; the curves for the two irradiated blocks are bracketed by the curves of the two non-irradiated blocks.

[0242] As described above, granulated carbon has been successfully used in dialysis, for example in the REDY, Biologic DT, and Allient machines in a non-sterile circuit. However, in some dialysis therapies, such as CVVHD (Continuous Veno Venous HemoDialysis), treatments are of long duration (up to 72 hours) and require a sterile dialysate circuit. In many cases, patient toxin (e.g., bilirubin) or ion (e.g., potassium) loads may be low or it may be deemed desirable to not remove beneficial substances from the patient (e.g., glucose). This may particularly be the case for patients suffering from drug overdose.

[0243] By way of an example treatment modality, FIG. 14 shows a simplified schematic of CVVHD. The dialyzer provides a bidirectional exchange of substances across the dialyzer for the purpose of equalizing concentrations of substances in the patient's bloodstream with the concentrations in the dialysate. This equalization is never perfect, but is a function of relative concentrations, time, and the permeability of the dialyzer membrane for a given substance. To move this equalization process forward, there typically must be continual replacement of the dialysate. Thus, the operation of the system is basically straightforward; fresh dialysate is simply pumped through the dialyzer into a waste container. To remove fluid from the patient (ultrafiltrate), the effluent pump pumps more fluid from the dialyzer than the replacement dialysate pump. The conventional method can require fairly high volumes of fluid and typically requires careful preparation of the replacement dialysate to assure sterility and purity. In this and succeeding figures, many required accessories such as blood leak and pressure monitors are not shown.

[0244] FIG. 15 shows how a carbon block can be used to purify the dialysate. This purification can take place both with respect to contaminants in the dialysis fluid as supplied, and also of substances removed from the patient. Due to the porous nature of the carbon block, air will be emitted from the block for some considerable time, so the arrangement shown or some other method will be necessary to prevent air from reaching the dialyzer. Such air does not normally enter the patient's bloodstream across the dialyzer membrane, but it does remove useful surface area from the dialyzer.

[0245] Those substances which are prescribed to be added to the patient may be loaded into the dialysate bag prior to the start of treatment. During a treatment, the carbon block, the dialysate bag, or both may be changed as needed.

[0246] As shown in FIG. 16, instead of, or in addition to, preloading the dialysate bag with prescribed substances, a separate infusion pump and infusate reservoir may be added in order to provide a continuous addition of substances to the patient.

[0247] FIG. 17 shows another addition; by adding an effluent pump and reservoir, there can be a continuous exchange of dialysate. In this case, the flow of dialysate from the infusate reservoir to the effluent reservoir can remove substances from the patient which are not well removed by the carbon block. (Note that in the former case, infusate will be delivered in relatively small amounts, while in this case, infusate will be used to exchange the dialysate in relatively large amounts.)

[0248] In certain other treatment methods, rather than adding infusate to the dialyzer, the infusate or other replacement fluid can be added directly to the patient's blood before the dialyzer, after the dialyzer or both. This may be done alone or in combination with any of the above described methods.

[0249] Although CVVHD is used as an example, the concept of using a sterile carbon block to regenerate all or part of the dialysate is applicable to a wide variety of therapies. Each of these therapies will have its own special plumbing arrangements; such different arrangements fall within the scope of this disclosure.

[0250] In regeneration of dialysate, carbon removes principally organic toxins greater than 100 m.w. This includes many middle molecules that have been shown to cause illness during kidney failure. However, there are some smaller and charged toxins of kidney failure that are not removed by carbon, including: urea, phosphate, sodium, potassium and acid. The acidity of blood is represented by a deficiency in concentration of various bases in the blood, and is corrected by addition of basic compounds such as bicarbonate. More complicated columns such as the Sorb include layers to remove these various small and charged toxins, but they require some careful management and priming to produce just the desired changes in body chemistry. With carbon regeneration of dialysate in CVVHD, the removal of larger molecular weight organic toxins can be greatly increased by merely increasing the dialysate flow rate. In standard CVVHD since the dialysate is sterile, pre-packaged and expensive, dialysate flow is typically 30-50 ml/min. This slow flow limits the chemical efficiency (clearance) of the system greatly. With carbon-regeneration of dialysate the flow rate can be increased to 400 ml/min without any increase in cost except the cost of the column. The removal of small charged toxins, and replenishment of bicarbonate can be simply provided by changing the bags of dialysate when required to supply the needed changes in body chemistry (such as several five liter bags per day). The concentration of the dialysate can also be chosen or adjusted for fine tuning the removal of the small, charged toxins. Thus, for the first time, CVVHD with charcoal regeneration of dialysate gives the physician the capability to control rate of removal of two different types of kidney failure toxins from patients, according to their needs: larger organic toxins and small charged toxins.

Section B: The Filtration Bed for Immobilizing Small Sorbent Particles.

[0251] Introduction The second technology for immobilizing powders which we have developed is a filtration bed which positions particles on the outside of the carbon block during fluid flow. For function in a hyperthermia circuit the sorbent used in the filtration bed will be calcium phosphate. The function of the calcium phosphate (CaHPO.sub.4) layer is to absorb one toxin (acid, H+) and to modulate or control levels of calcium and phosphate in the dialysate. Working through its solubility product, calcium phosphate will release calcium or phosphate if their levels are low in the dialysate. If the levels are high, it will remove calcium and phosphate. The dissolution or creation of calcium phosphate is possible only when there is a very high surface area/weight, meaning very small particle size (such as a few microns). When employed in the original BioLogic-HDT circuit, the calcium phosphate in the dialysate was precipitated on the surface of the carbon powder particles and held in suspension. The suspension moved through the dialyzer, propelled by membrane motion and vacuum/pressure gradients. In the current application, the calcium phosphate will be a powder that is held motionless in a filtration bed around the carbon block. Other applications of sorbents also require very small particle size, such as use of microporous crystals of zirconium silicate, for binding potassium and ammonium in a dialysate circuit). Note that calcium phosphate is exemplary; other substances may also be used.

[0252] At a modest flow rate such as 250 ml/min a finely powdered sorbent, such as calcium phosphate (CaHPO.sub.4) will form a layer fixed on the outside of the carbon block. Fluid flow through the layer proceeds without any significant pressure gradient (with 100 grams of calcium phosphate, about 60 mm Hg pressure drop). With perfusion of dialysate around the particles, calcium phosphate powder can dissociate and deliver soluble phosphate whenever the dialysate calcium x phosphate product decreases below the dissociation constant for calcium phosphate, just as it did in the suspension of the BioLogic-HDT system. The photographs below (FIG. 18) show the carbon block and calcium phosphate powder without fluid flow through the carbon block (left) and with fluid flow of 400 ml/min (right). With fluid flow, the calcium phosphate powder is firmly applied to the outside of the carbon block, but fluid flow continues through this filtration bed of particles without any significant increase in pressure gradient (57 mm Hg at 400 ml/min flow rate). When flow is stopped, the calcium phosphate powder falls downward to the bottom of the canister (as shown on left), and the powder will re-suspend and apply itself to the outside of the carbon block when flow resumes. None of the powdered calcium phosphate penetrates into the block (as shown by sections of the block after use), and no particles permeate the block. The calcium phosphate powder is in intimate contact with all fluid flowing through the filtration bed and apparently, the fluid flow is very uniform.

[0253] Concept of Structure and Function of the Carbon Block/Filtration Bed, and Why Flow and Function is Different from a Standard Sorbent Column

[0254] It is helpful to compare the present invention with standard packed columns. With a standard sorbent column, large particles or granules are used as sorbent. As discussed above, the finest particles used within sorbent columns is approximately 50 microns, and this small size allows uniformly distributed flow without very high pressures only if the particles are spherical. For applications of carbon in columns the particle size is usually quite large (such as 1-2 mm) and the individual granules are easily palpable. To load a standard column, the dry granules or sorbent particles are usually poured into the open column, a top is attached, the column is inverted to begin filling (allowing air to escape) and fluid flow is begun. When the air has been expelled, the column is inverted again. Sometimes the column is filled with fluid and then the sorbent particles are poured in. Whether filling a wet or dry cannister, the force of gravity and chance determine the position of granules when perfusion starts. Larger granules are interspersed with smaller ones. If one area has a greater proportion of large granules or a small channel space, then during fluid flow through the column this channel will widen and fluid flow here will be more rapid than that through the rest of the portions of the column, as can be demonstrated during dye injection. The result of this rapid flow is early saturation of the sorbent granules of the channel, and subsequent early breakthrough of bound toxins or compounds. Further, the interspersing of large and small granules tends to form a tight pack (much like occurs with use of varying gravel size used in road construction). However, to a large degree, the use of uniformly sized particles, sophisticated column packing techniques, packing fluids and apparatus can greatly reduce these problems. Such techniques do, however add significant cost to the column. When we have attempted to make a column out of our calcium phosphate powder, we have found that when we begin fluid flow the powder forms a very dense semi-solid cake and perfusion pressures at low flow rates are in the hundreds of mm Hg, for columns that are only about 1 cm thick.

[0255] The method of constructing the outer powdered sorbent layer of the carbon block/filtration bed device is quite different. The loose and very fine powder is placed in the bottom of the canister, and the fluid flow is begun. The fluid flow rate exceeds the sedimentation rate of all of the particles and therefore the particles are carried with the fluid against the force of gravity. It is likely that during the fluid flow the finest particles are carried to the surface of the carbon block first, then the larger granules. As the particles form layers around the various portions of the carbon block, then hydraulic resistance of each portion becomes higher and fluid flow automatically re-directs to portions that do not have a powdered bed layer. After the entire carbon block outer surface is covered, then there are probably still portions which have higher flow. However, the higher flow in these channels brings with it more sorbent particles and the channels tend to fill and resolve automatically. The powdered bed appears to be less likely to pack tightly compared to a standard column. Whereas it required several hundred mm Hg of pressure to perfuse a standard column created from calcium phosphate powder, the carbon block/filtration bed held about 50 grams of calcium phosphate powder and when perfused at a rapid rate of 250 ml/min had a pressure gradient of about 57 mm Hg.

[0256] Another feature of the carbon block/filtration bed that distinguishes it from a standard column is the shape of the sorbent layer. Instead of being a cylinder with fluid flow along its axis and a filter at one end, the filtration bed on the carbon block forms as a layer around a cylinder with fluid flow on an inward direction normal to the surface of the cylinder. The use of the outer surface of the carbon block as the filtering surface means that there is a very large surface area for filtration and support of the powdered sorbent bed. This means that a very large amount of powder may be applied to the surface of the carbon block without creating a thick layer of powder. As an example, one size of the Matrixx KX-5 is 2.5 inches in outer diameter and 10 inches long. The circumference is thus about 8 inches and the surface area of the outer portion is about 80 square inches. If this same surface area were created as a flat filter at the bottom of a cylindrical column, the diameter would be approximately 10 inches (about 25 cm). If the desired thickness of the sorbent layer were only 1 cm, this would result in an aspect ratio (width:height) of 25:1 for the column, a configuration which would certainly encourage irregular flow. However, with the filtration bed, it appears that flow is uniform through all parts of the bed (judging from the structure of the bed alone). If a standard column were created with a more standard aspect ratio such as 1;1 or less, then to utilize the same amount of powdered sorbent it would require a column height many times higher. The longer fluid flow path would greatly increase the hydraulic resistance of the column. The large surface area of the outside of a cylinder has a second advantage, in that it diminishes the rate of fluid flux through the sorbent layer (flow rate per cm2 of filter surface). The result is increased dwell time which improves reaction kinetics. This decreased flow rate also decreases the hydraulic pressure drop through each cm2 of sorbent bed. This benefit is of course another way to describe the benefits of a very high aspect ratio for the filtration bed. A simple depiction of differences between a standard column and the carbon block/filtration bed approach is shown in FIG. 19a.

[0257] Clearly, other variations on this same principle are possible. For example, a column may be constructed with or without the carbon, having a membrane, filter, screen or other means (designated screen hereafter in this paragraph) with which to constrain particles. Multiple geometries are possible. In all cases, there are three requirements. First, the zero-flow position of the particles must be substantially away from the screen. Gravity would be the normal means of achieving this, but reverse flow is also a means. Secondly, the particles must readily suspend during flow by means of an appropriate combination of particle size, fluid density and viscosity, other fluid characteristics, fluid/particle affinity, surface tension, etc. Thirdly, the particles must have limited affinity for one another to avoid clumping and other undesirable aggregation. Surfactants in the fluid may possibly be included in the fluid to aid in meeting these requirements. FIG. 19b exemplifies this concept. Also, of course, carbon or other materials or sorbents may be formed into solid porous blocks (in place of the screen of FIG. 19b) of various shapes by which means fluid volume and space requirements may be reduced for a given surface area. A vertical system is also quite possible; the screen is at the top of a short column of large diameter.

Use of Carbon Block/Filtration Bed to Regenerate Dialysate in a Dialysis Machine

[0258] With the carbon block and filtration bed of calcium phosphate (plus the cone reactor as described below, in some circumstances) we can recreate the chemical function of the BioLogic-HDT system using a dialysate regenerating system in which dialysate flows uni-directionally through the canister. This system is more conventional than was the sorbent suspension system, is more similar to a standard sorbent column, and is easily compatible with regeneration of dialysate flowing through a standard hollow fiber dialyzer. The powdered carbon will effectively remove almost all organic toxins which penetrate the membranes. The calcium phosphate will operate by solubility product to modulate the dialysate concentration of calcium, phosphate and bicarbonate. When any of these electrolytes become abnormally low, the calcium phosphate will automatically replenish them. When any of these electrolytes become abnormally high, the calcium phosphate will remove them. The Dialysis Machine

[0259] For treatment of patients in the current protocol we will use the carbon block/filtration bed canister for removal of toxins from dialysate and provision of phosphate whenever dialysate phosphate diminishes below normal. The carbon block/filtration bed will be provided in clean form and incorporated into the dialysate side of a standard NxStage System 100 dialysis system. The NxStage System 100 is a commercially available high permeability dialysis system that is used in many hospitals for continuous dialysis of patients in the ICU. It is also used in treatment of home hemodialysis patients, usually on a short daily schedule. The NxStage machine controls ultrafiltration (UF) automatically through use of two dialysate side pumps, two volumetric chambers and an ultrafiltration pump. The NxStage system is used in the hospital setting with pre-mixed 5 liter bags of sterile dialysate (bicarbonate based). At home, it is often used with a 60 liter bag of lactate based dialysate created on site with the PureFlow device. Maximum blood flow rate is 500 ml/min and maximum dialysate flow rate is 250 ml/min. In hyperthermic therapy the NxStage dialysis system will be connected in parallel to part of the blood heating circuit, similar to how the BioLogic-HDT was connected in parallel to the blood heating circuit in the previous BioLogic-HT System. However, with the new system we will control blood flow rate through the dialyzer with the blood side roller pump of the NxStage device, at a controlled rate of 400 ml/min. We will therefore be able to remove blood after the roller pump and replace it just before the heat exchanger, in a co-current mode with all other blood flow in the HTA portion. In the BioLogic-HDT system blood flow was passive through the dialyzer, and counter-current to all the other blood flow, creating significant recirculation of blood through the dialyzer. This recirculation is avoided with the Generation II system.

ThermalCore-HT Circuit Schematic

[0260] Blood flow through the heating circuit will be from 1000-2500 ml/min. The following, FIG. 20, is a diagram of the heating circuit of the ThermalCore HT System which includes the DeBakey roller pump and BioTherm heat exchanger, and the NxStage sorbent-dialysis system.

3.4 the ThermalCore Ht System Operation

[0261] With the NxStage system as the HDT circuit we will use 5 liters of bicarbonate-based dialysis fluid. This will provide a larger amount of potassium and bicarbonate than was present in the two liters of fluid in the original HDT circuit, and a greater volume of dialysate for removal of potassium if needed (by using a low potassium concentration in dialysate). The total capacity for balancing electrolytes should be essentially the same as was present with the original HDT circuit containing the electrolyte-balanced polystyrene sulfonate (which remained mostly loaded with divalent cations calcium and magnesium). Changes in calcium, phosphate and bicarbonate concentration will be offset through dissolution of precipitated calcium phosphate (powder), as it was in the original HDT system. We will circulate the dialysate at 250 ml/min, through the dialyzer, through the charcoal block/filtration bed canister, and back to the bag. We will not need a heater in the NxStage circuit, as the 5 liters of dialysate should quickly come to nearly the same temperature as the blood within the patient. We expect to set the ultrafiltration rate of the NxStage circuit to zero, but if it appears the patient has received more fluid than needed, UF could be removed at up to 1000 ml/hour. This ultrafiltered fluid would accumulate in the 5 liter bag, which is used to prime the entire dialysate side of the circuit.

[0262] With incorporation of the NxStage System into the ThermalCore HT system, we are using a commercially available and well-proven device to automate the dialysis circuit, monitor ease of blood flow, detect bubbles, control ultrafiltration, and limit blood side chemical changes.

[0263] These features and functions are all similar to those that were included in the BioLogic-HT System, but we accomplish these functions using technology that appears much more conventional. The many similarities in function between the original BioLogic-HT System and the current system are demonstrated by the following Comparison Table:

TABLE-US-00001 TABLE # 16 Comparison Table of the Original BioLogic-HDT System and the ThermalCore-HDT portions of the Systems: Feature BioLogic-HT ThermalCore-HT Dialyzer Cellulosic Flat Plate, Polysulfone hollow fiber, 1.8M.sup.2 1.6M.sup.2 Blood Flow Rate 600-800 ml/min with 400 ml/min without recirculation recirculation Dialysate Flow Rate 300 ml/min (net out 250 ml/min of dialyzer) unidirectional Creatinine clearance 130 ml/min 150 ml/min (in vitro) Ultrafiltration Rate 0-1000 ml/hour Same Powdered Activated 140 grams, Coconut, 300 grams, Coconut, Charcoal in suspension supported in carbon block Powdered Calcium 50 grams (80 mM), 50 grams, precipitated by Phosphate USP precipitated in bag manufacturer Potassium removal 10 meq 60 meq (from one 5 liter maximum (with bag, a second bag could be zero potassium used to contribute more) added to bath, patient K of 6) Potassium donation 6 meq Same maximum (starting bath of 6 meq, patient K of 3) Bicarbonate 40 meq Same donation maximum (from 2 liter bag) (patient bicarbonate of 10) Phosphate donation 20 mM Same maximum (patient phos of 0.5 mM) Bacteriologic Status Clean, not sterile Same of dialysate circuit Blood Temperature Outflow of Heater, Same Monitoring Inflow Blood Line Patient Temperature Multiple Points Same Monitoring

[0264] In terms of function and features, steps of operation, and clinical effects we expect the ThermalCore-HDT System to be highly similar to that used in our prior IDE studies. However, the overall operation will be much simpler.

[0265] There are other potential uses for the carbon block/filtration bed technology besides whole body hyperthermia. If the calcium phosphate powder is replaced by an ammonium sorbent such as a cation exchanger like powdered microporous fractionated protonated zirconium silicate (ZS, U.S. Pat. Nos. 5,891,417, 6,579,460, 6,099,737, and published application 2004/0105895), then the carbon block/filtration bed technology should be perfect for treatment of liver failure. If the urease enzyme is bound to the ZS or placed in a layer upstream from it, then the system could effectively treat kidney failure (an anion exchanger would also be needed). If an immune-sorbent is used in the filtration bed and the perfusate is plasma, then various immune diseases might be treated such as lupus erythematosus, Wegener's, rheumatoid arthritis and psoriasis. The charcoal also will bind a number of intermediaries of these immune diseases. Finally, with a sorbent capable of binding endotoxin and TNF (a cytokine) such a system with plasma perfusion could treat the condition of sepsis.

[0266] If there is one down-side of the filtration bed, it is that when fluid flow is stopped, the sorbent particles leave the membranes and quickly fall to the bottom of the canister. When fluid flow is re-started there will be some passage of the toxin materials from the bulk fluid through the carbon block, until the filtration bed is re-established by the flow. For toxins of low potency to the patient, this is not a problem. For some toxins such as ammonium created by urease, release to the patient could cause problems. If this is a problem, then there are ways to maintain fluid flow through the filtration bed when blood flow through the dialyzer is stopped. The easiest is to merely continue dialysate flow, even if blood flow is ceased through the dialyzer. Dialysate flow could be bypassed around the dialyzer if such is required.

[0267] We have also found that to form a fluidized bed of small particles the flow rate through the CB must be relatively high, such as 400 ml/min for a CB of 2.5 diameter and 10 length. At 250 ml/min the fluidized bed does not form without agitation of the CB and suspension. Further, the formation of the fluidized bed depends on the particle size of the suspension and the density of the particles. For particles over 10 microns in size of reasonably high density such as over 2 gm/cm3, and relatively low flow rate, the fluidized bed may not form well at all. If the fluidized bed does form but becomes too thick, then flow through the CB is very irregular.

[0268] There are also many other cases in which, in contrast to the discussion surrounding FIG. 18, powder will not naturally uniformly coat a carbon block, but will form an uneven layer. By using a cone-created fluidized bed, very fine particles which are not contained by the cone reactor will form a uniform layer on the carbon block. As a result, all particles will either be suspended in the fluidized bed or uniformly coat the carbon block. This uniform coating naturally occurs because only freely floating particles will reach the carbon block and uniform flow through the block will uniformly distribute them.

[0269] For all of these reasons we have decided to combine the CB and FB with a conical reactor containing a fluidized bed. The fluidized bed will work to perfuse fluid through particles that have too high a sedimentation rate to rise and form a fluidized bed around the CB. Those particles with smaller size will rise and form a layer as FB around the CB, as described below. This layer will be relatively thin and made of small particles, and results in a uniform fluid distribution through the CB.

Section C: Cone Reactor with Fluidized Bed for Use in Combination with CB and FB Introduction

[0270] The cone shaped reactor is a device to contain a fluidized bed, containing all particles in a suspension with sufficient density and particle size to remain in the reactor during upward flow of fluid. Initial experiments showed that the sorbent calcium phosphate (CP) forms a cloud of relatively dense particles, topped by an area of finer particles. It was immediately realized that a cone shaped reactor could permit an equilibrium between the linear flow velocity of the fluid and the settling rate of particles and also allow the fluidized bed to continue to operate over a range of fluid flow rates.

[0271] Initial experiments with an Imhoff cone (similar to a funnel with sides 7 off vertical) confirmed this hypothesis. In this experiment, the CP was placed on top of a frit made of a piece of porous plastic with 35 m nominal pore size. The cone was provided with a lid with which to return fluid to the reservoir. It was found that when fines released by the cone returned to the cone, they gradually plugged up the inlet frit and pressure built up unacceptably. This experiment did, however confirm the basic principle of the cone reactor concept.

[0272] At the suggestion of David Carr, a carbon block in an un-modified filter holder was used to catch the fines. This method worked well for both anhydrous particle sizes and for dihydrous CP. The main findings are summarized in Table 1.

TABLE-US-00002 TABLE 1 Fluid Velocity vs. CP Bed Height Fluid Velocity CP Flow Cloud Cone at Top of Cloud Type Rate Height Angle (cm/min) Old Anhydrous 250 21.5 7 3.8 Old Anhydrous 100 14 7 3.6 New Anhydrous 117 21.5 7 1.78 New Dihydrous 86 21.5 7 1.31

[0273] Also, to determine the limits of cone angles, both anhydrous and dihydrous CP were poured into funnels of various angles, including angles beyond that of the bare funnel by mounting the funnel in a ringstand and tilting the ringstand. The CP was allowed to settle and the funnel surface was examined. Then the funnel was drained at a flow rate determined by gravity and 4.5 mm ID tubing. The funnel surface was then examined again for residual powder. It was found that funnel angles up to 45 degrees off vertical were tolerated, with only a minor dusting of powder on them.

[0274] As a result of these experiments, a spreadsheet was created to assist in the analysis and design of cone reactors for CP. To test the validity of the spreadsheet and the functionality of the combination of the cone reactor and carbon block with a coating of CP powder, 3 experiments were performed. We shall call the combination of carbon block with a bed of CP on it (CBFB) and a cone reactor a CCS (CBFB plus Cone reactor System). The output of the spreadsheet is shown in FIGS. 46 and 47.

[0275] Analysis also revealed that while a pure cone reactor would work, the large volume of a cone as one goes up in diameter results in fairly useless and large extra volume. Hence, a more volume-efficient cone reactor uses a cylinder on top of a cone, roughly in shape to an unfritted Buchner funnel. That said however, it was found (see Results) that 1-2 cm of headspace in the cone prior to the cylinder seems to reduce fines emission from the effluent. In fact, an overloaded cone will naturally have a level 1-2 cm below the cylinder.

[0276] With this information in hand, three experiments were performed.

[0277] As may be seen in FIG. 21, following the arrows from the reservoir on the left, fluid is pumped using a roller pump to the bottom of the cone reactor. The effluent from the cone reactor goes to the filter holder, thence to the outside of the carbon in the CBFB, then through the carbon block to the center hole and out back to the reservoir. The reservoir was heated to 41+/1 C., and stirred continuously. Flow and pressures across the two reactors were acquired by a proprietary data acquisition system.

[0278] The reactor was an ordinary laboratory funnel with sides a 30 angle from vertical. A cylinder 15.2 cm inside diameter was placed on top of the funnel and sealed with permanently sticky butyl caulk. A top was provided for the cylinder with an O-ring and provision to adjust the height of the top so as to vary the cylinder height and volume. Calculated volumes for the cone reactor were 793 mL for the cone and 905 mL for the cylinder giving a total of 2968 mL for a 5 cm headspace. Each additional cm is approximately 181 mL.

[0279] In each experiment, flow was initially set to approximately 250 mL/min. Conditions were varied as seemed necessary or as thought might yield interesting results. The first experiment was designed to test a steady state condition. The second experiment had more of a goal to break the functionality of the CCS, and the third experiment was designed to test the interface with the NxStage machine.

[0280] 250 mL/min was selected as it was thought that this is the maximum NxStage flow; the maximum is actually 200 mL/min.

TABLE-US-00003 TABLE 2 Experiment Conditions Summary Initial Anhydrous Initial Experiment CP Load (g) Flow (mL/min) Headspace (cm) Fluid Remarks 1 50 250 9.5 0.9% NaCl 250 mL/min entire run except for short run at 100 mL/min and stop-flow test. 2 50 262 5.0 RFP-404 370 mL/min tested without ill effect. Also tested at 100 and 160 g without ill effect. Fluid was at room temperature (20 C.) until t = 33 min, then heater turned on, set to 42 C. 3 33 200 5.0 RFP-404 + NaCl NxStage Machine testing. Machine had pauses which collapsed CBFB bed.

[0281] In FIG. 28, note how there is a line just below the rim of sealant. That line is the start of the cylinder. The check valve may just be seen above the worm drive clamp at the bottom. It is the black ring inside. The check valve was made of half of a rubber stopper, top diameter 13.1 mm, bottom diameter 10.9 mm, length 14.2 mm, with a long screw to weight it down and keep it straight. Total mass was 3.54 g. The check valve was not intended to stop reverse fluid flow, only keep powder from exiting the cone reactor during flow stop, a job it did well. In experiments omitting the check valve, powder consistently entered the influent tubing at zero flow. It should also be possible to use a clamp around the tubing if the screw extends below the cone into the tubing. The temporary clamp retains powder during shipping. A similar scheme can be used at the outlet. In production units, other refinements are possible such as using a plastic rod instead of a screw and scoring the plastic rod whose end is attached to the tube. The user bends the tube to break the rod at the score to release the check valve for operation.

[0282] Observe in FIG. 30 how much lower the suspended CP bed is. Much of the CP has gone to the CBFB. CP not retained by the cone reactor was deposited on the carbon in the CBFB as seen in FIGS. 31 and 32.

[0283] FIG. 32 shows that no harm had come to the CBFB's flow uniformity, but during a stop-flow test, some CP was not retained on the carbon and fell to the bottom.

[0284] Significant observations included the ability of the CCS to perform well at 100 mL/min, and partly re-start after a stopped-flow condition.

[0285] At the end of experiment 1, the CP in both the CR (Cone Reactor) and CBFB were washed into beakers. The supernatant was sucked off after an overnight settling time, then the remaining substance dried. The CBFB was found to contain 12.5 g and the CR 32.8 g, including a 5 g loss. Thus, the CBFB ended up with 28% of the CP by weight.

[0286] The natural cloud height was about 1.6 to 2.2 cm below the start of the cylinder. This is not a hard and fast ruledue to the stochastic processes involved, there is never an actual cessation of particle carryover to the CBFB.

Experiment 2

[0287] As may be seen in FIG. 33, the fluid volume of the CBFB is not more than 935 mL. The weight steadily increases; the proportions due to powder accumulation and air emission are not known.

[0288] Significant observations in Experiment 2 included: [0289] Change in salt solution from NaCl to RFP-404 had no effect [0290] Change in startup temperature from 42 C. (Experiment 1) to 20 C. (Experiment 2) had no noticeable effect. [0291] CP could be slurry loaded, but larger particles remained in pump tubing for the duration of the experiment. [0292] The ability to successfully slurry load an additional 50 g at t=60 minutes for a total of 100 g and at another 60 g at t=111 minutes for a total of 160 g, with an increase in cloud size and particle carryover to the CBFB, but without any drastic effects. [0293] The ability to operate at 50% greater than design flow rate with only additional particle carryover. [0294] In the event of an overload, where the cloud extends into the cylinder, the cylinder essentially becomes a particle classifier as seen in FIG. 36. [0295] Worst case CBFB powder load did not impair uniform flow through the carbon. [0296] Increasing CP load does significantly increase cloud size. [0297] The final CP mass in the CBFB was 28.19 g, and in the CR 131.4 g. Loss was less than 0.5 g. The percent of CP in the CBFB was 18%, and in the CR 82%.

Experiment 3NxStage Machine

[0298] This experiment was primarily designed to test the interface with the NxStage machine. One objective was to simply test overall functionalitywould the two systems work together with a significant safety margin.

[0299] In FIG. 38, The box denoted by the red arrow includes a CR followed by a CBFB. The MCH-1000 and Biotherm were simulated by a simple resistance of about 80-120 mmHg. Flow through the CCS was 200 mL/min. Flow through the simulated Biotherm/MCH-1000 was typically about 754 mL/min.

[0300] To test the ability of the NxStage machine to tolerate the CCS with a good safety margin, three tests were performed, which, along with periodic machine starts and stops will explain FIGS. 39, 40 and 41.

[0301] In the first test, an adjustable flow restrictor was placed in the CBFB effluent line. The effects may be seen at t=50 to t=60 in FIG. 40. The NxStage machine did not give any alarms or error indications. The approximate 300 mmHg limit was the operator's comfort limit.

[0302] In the second test, note in FIG. 38 that both the venous and arterial lines are essentially at the same pressure. A flow restrictor was placed at the outlet of the blood line (returned to the patient). The effects may be seen from t=60 to t=68 in FIG. 41. The machine generated alarms at 125, 156 and 143 mmHg. The reader is cautioned that the machine algorithms with respect to rise and fall times, time delays and limits are not known and should be determined from the manufacturer.

[0303] In the third test, the main blood flow pump was suddenly turned off or on, as may be seen at t=69 to t=75. In all cases, the pump could be suddenly started without a response from the machine. However, if the pump was suddenly stopped the machine would generate a non-fatal alarm. The pump could be stopped by a ramp-down over a period of very approximately 3-5 s without generating an alarm.

[0304] The NxStage also never complained about the temperature of the blood. See FIG. 42.

[0305] Three times however during the final 115 minutes of the experiment, the NxStage machine stopped dialysate flow. The duration and spacing of these pauses is apparently random. During the final pause, the CP bed on the CBFB collapsed. At no time was uniformity of flow impaired, but unused CP was left at the bottom of the filter holder.

[0306] The CP level in the funnel was noticeably lower than in earlier experiments. Unfortunately, the measurements are lost due to a change in the funnel configuration that was overlooked.

[0307] NxStage Operating Notes [0308] The NxStage machine has a maximum DFR of 200 mL/min (recently increased to 400 ml/min). [0309] The return from the CCS needs to go to a separate bag port than the main port to avoid adjacent lines ingesting air. It would also appear helpful to raise this port slightly to keep air well away from the main port. This will require a dual-male luer connector not provided with the disposables kit from NxStage. [0310] It is necessary to break the frangible seal on the extra bag port. [0311] The line with the green clamp needs to be connected to the bag.

Recommendation

[0312] FIG. 43 (not to scale) is a schematic of a significant refinement of an idea suggested by Dr. Steve Ash. The issue with pauses can be eliminated if the falling CP is returned to the cone reactor. Additionally, the number of modifications to the original CBFB system have become sufficiently extensive as to warrant a fully custom unit. In FIG. 43, the inlet check valve is not shown. The cone plus cylinder version of the cone reactor is used. Directly on top of it is a carbon block to create a CBFB. The lower, inner hole of the carbon block is plugged. Flow begins at the bottom, proceeds to the carbon area, goes through the carbon into the inner hole in the carbon and out the top outlet. It may be found expedient to fill the bottom lip of the carbon filter end cap to keep it from trapping falling CP. Alternatively, a block without an end cap may be used and a blank plate substituted.

[0313] Detailed design of an actual CCS as shown in FIG. 43 must await one or two additional experiments and work authorization.

CONCLUSIONS AND SUMMARY

[0314] Definition: CCS a combination of a CBFB+a Cone Reactor System.

[0315] Evacuating the CBFB is may be contraindicated for some applications unless it is filled off-line. The CBFB may thus be either filled off-line or shipped filled with liquid. In the latter case, the CBFB needs to be well shaken before use.

[0316] A CCS, or CBFB may possibly need to be placed on the floor or a table to avoid disturbance due to machine vibrations. Excessive vibrations from some machines could disturb the filter bed.

[0317] Cone reactor hydraulic resistance is negligible, about 10 mmHg at 250 mL/min. Fittings will do that. (Note that a 31 inch height difference in water levels produces about 60 mmHg pressure differential.)

[0318] There should not be a distributor in a cone reactor. A check valve consisting of an elastomeric plug with a long rod extending into the inlet tube acts as a weight and retainer. This simple method keeps powder out of the lines during shutdown. A clamp on a tube can hold the rod tight against the funnel inlet to keep powder contained during shipping. Other methods are possible as well.

[0319] Effective dwell time is some fraction, around 0.8 of cone volume, not entire reactor volume, divided by flow.

[0320] A cone reactor performs well at any flow which places the cloud top less than about 2 cm of top of cone. In this case, the filter bed will be relatively thin.

[0321] Higher flows push the cloud into cylinder. The cylinder then acts as a classifier and retains larger particles while sending smaller ones to CBFB. In this case, the filter bead will be relatively thick.

[0322] A good length of time, about 3 h is required to fully evaluate a CCS. This is due to the statistical nature of fluid and particle flow. This does not imply that the adsorption kinetics of the powder change significantly during this startup time.

[0323] Particle transfer is a function of fluid mechanical and thermodynamic probabilities, as well as the particle size distribution. The situation is analogous to water in a pan on the stove. Eventually it will evaporate. Turn on the gas and it will boil. But the process is essentially the same. Thus, particles are continuously, at some rate, transferred from the CR to the CBFB.

[0324] Cloud volume, to a large degree, is a function of the amount of CP loaded into the system.

[0325] At end of experiment 1, 50 g of CP put 13 g on the CBFB, leaving the rest (72%) in the CR. (12.5 g/32.8 g=28% in CBFB, 72% in CR5 g lost). In the second experiment, 18% was in the CBFB and 82% in the CR. In the third experiment 12.5 g was on carbon, and 34.6 g in CR, for 27% and 73% respectively.

[0326] Since the CBFB receives the smallest particles from the CR at a slow rate, the buildup on the CBFB is uniform. Flow is uniform through the block.

[0327] Due to the stochastic nature of particle retention and transfer, overloads of the CCS, whether from excessive flow rate or excessive CP load, smoothly transfer CP mass to the CBFB at an increasing rate without sudden breakdowns of the process. This assumes that nothing is grossly undersized.

[0328] Depending upon schedule, it is possible that further refinements may be made in the design. E.g., it is not known at what point the CBFB becomes actually overloaded and fails to provide uniform distribution.

[0329] CBFB resistance goes up initially, then down with time. This is likely to be a function of fluid temperature, possibly from changes in fluid density or increased Brownian motion. Particle dissolution could also be a factor.

[0330] The 9 to 11 m particle size Anhydrous CP seems to have a fairly wide distribution. Some particles remained in pump tubing during slurry fill.

[0331] Under some conditions, slurry fill of powder may be possible with very fine powders. This may be useful if a powder must be added to an already running system.

[0332] A cone reactor may be restarted after 25 minutes.

[0333] A CBFB with a light coating will start falling off after about 25 minutes after a stop-flow event. Data suggests that hold-up time may be a function of trapped air continuing very small flow for a time.

[0334] Comparing dwell time to dissolution data indicates reasonable feasibility.

[0335] Crude estimate of CP needed: 250 mL/min*60 min/h*3 h/1000=45 L. 45 L*0.2 g/L=9 g. Another source gives 0.316 g/L at 38 C., for 14.2 g. Using 50 g given to FDA, corrected for molar ratios gives a dihydrous load of 63.5 g, which should be enough.

[0336] The combined CCS of FIG. 43 has significant advantages and may be easily manufactured by any competent machinist in prototype quantities or in large quantities by normal plastic injection or blow molding.

Section D: Prior Art Use of a Suspension Powdered Sorbents for Dialysate Regeration, the BioLogic Series of Devices

[0337] The following is a description of prior art for using powdered sorbents to regenerate dialysate in an extracorporeal circuit. The prior device was the BioLogic-HDT system, with a suspension of powdered sorbents which passed directly through a plate cellulosic dialyzer.

[0338] The BioLogic-HDT System

[0339] An important advance in whole body hyperthermia came with the use of a dialysis system with sorbents to remove various organic toxins and to limit changes in various electrolytes such as phosphorus and bicarbonate during treatment. This was accomplished in our prior studies of PISH through use of an adaptation of the BioLogic-DT (Liver Dialysis) machine. The DT machine had a suspension containing activated charcoal powder to remove organic toxins, and a sodium-loaded cation exchanger powder (polystyrene sulfonate, PSS) to remove small amounts of potassium and ammonium. In the HDT system we also precipitated 80 mM of calcium phosphate within the sorbent suspension (50 grams). Through dissolution, the calcium phosphate precipitate would increase the phosphate level in dialysate if it was lowered by a decreasing phosphate concentration in the blood. We also changed the loading of the cation exchanger so that it was in equilibrium with the normal plasma levels of potassium, magnesium, calcium and hydrogen. In this way, the PSS would release a small amount of these cations into dialysate if concentrations fell below normal, and absorb a small amount if the concentrations rose above normal. The automatic control of changes in dialysate chemistry would then offset and diminish changes in chemical concentration in the blood during whole body hyperthermia treatment (WBHT) therapy (called perfusion-induced systemic hyperthermia (PISH) if perfusion-induced).

Description of the Biologic-HDT Machine

[0340] The following is our description of the Biologic-HDT dialysis circuit from our 1996 Food and Drug Administration Investigational Device Exemption (IDE) (G960257/S) for a clinical trial of WBHT in treatment of patients with advanced lung carcinoma (page 6 of the Operator's Manual):

[0341] The dialysate side contains a 2-liter suspension of powdered sorbents (charcoal and cation exchanger) which circulates between the dialyzer and a bag, and controls chemical composition of the dialysate according to binding characteristics and loading of the sorbents. The ultrafiltration rate is measured by changes in weight of the sorbent bag; simple algorithms adjust the ratio of blood inflow and outflow cycle times to obtain the minimal ultrafiltration rate, and automatically reinfuse fluid to the patient to obtain exactly the prescribed weight increase or decrease during treatment. The sorbent components of the BioLogic-HDT machine were as follows: [0342] 1. 200 grams of cation exchanger, pre-loaded with sodium, calcium, potassium, hydrogen, and magnesium in amounts to maintain equilibrium with normal blood concentrations, [0343] 2. 140 grams of powdered activated charcoal with 80 mMoles of calcium phosphate (50 grams) precipitated on the surface to dissociate whenever the surrounding solubility product is lower than the normal blood solubility product, [0344] 3. Sodium bicarbonate and sodium chloride in physiologic concentrations, [0345] 4. Flow-inducing agents, [0346] 5. And glucose absorbed to the powdered charcoal to dissociate and maintain normal or slightly high blood glucose.

[0347] Additional calcium chloride is infused into the venous return line of the HT at a rate necessary to maintain a normal blood calcium concentration. In-room analysis of plasma phosphate prompts addition of disodium phosphate to the sorbent suspension whenever plasma phosphate decreased below normal levels (which happens only during high temperature WBHT). From previous studies, a solution of disodium phosphate and sodium bicarbonate was devised which, when added to the sorbent suspension during high-temperature PISH, should result in normal blood chemistries at the end of the procedure, eventually obviating the need for performing blood chemical analysis during PISH.

[0348] The BioLogic-HDT system contained a plate dialyzer in which membrane expansion and compression mixed the sorbent suspension at the membrane surface. In the BioLogic-HDT system blood through the dialyzer was passive, created by positive pressure on the return limb of the blood circuit, and carrying blood back to the inflow limb where pressures were negative. Blood flow rates were 1500-2000 ml/min through the roller pump/heating circuit and 600-800 ml/min through the dialyzer. Blood flowing through the HDT portion returned to the inflow side of the roller pump, so there was some recirculation of treated blood through the dialysis system. Sorbent was circulated by alternating pressure in a reservoir on the outflow side of the dialyzer. The following diagram of the circuit was included in our 1996 IDE Application:

BioLogic-HT Circuit Schematic

[0349] The BioLogic-HT Circuit is shown in FIG. 44.

[0350] Results of Clinical Trial with the BioLogic-HT system

[0351] Clinical trials of the BioLogic-HT system demonstrated that during PISH with this system there were minimal changes in calcium, magnesium, phosphate and serum bicarbonate. Further, the patients remained physiologically stable with modest fluid replacement, during the WBHT treatments.

[0352] After initiation of our clinical trials of the BioLogic-HT system in treatment of patients with cancer we received FDA approval to market the BioLogic-DT system for treatment of hepatic failure with coma or drug overdose. Initial marketing efforts of this treatment were highly successful, but wider market entry was limited by the need for a specialized machine for this therapy, requiring installation of a new machine and training at each hospital planning to treat patients with liver failure or drug overdose. Currently the BioLogic-DT system is no longer available through its manufacturer, and the plate dialyzer is no longer available. The BioLogic-DT system with some modifications was the device used in the BioLogic-HT System.

[0353] In terms of other prior art, another method for constraining powdered sorbents to allow perfusion is a nanofiber felt. If layers of nanofiber polymeric materials are bound to powdered sorbents and then either rolled up or layered, the fine powder particles are held motionless during perfusion. There are almost no fines released during perfusion and flow distribution is good. The downside is that there is a very low packing density. Only about 10% of the volume of the nanofiber felt layers is due to the sorbent particles. By comparison, the carbon block and filtration bed are each more than 80% by weight and 50% by volume of powdered sorbent.

[0354] Another technology we developed for powdered sorbent regeneration of biologic fluids was to create a bidirectional flow of plasma from blood through membranes, allowing the filtrate to contact powdered sorbents in a suspension transiently and then return to the blood. This application was implemented in the BioLogic-PF for plasma depuration, and also was shown to work with hemofiltration membranes (membranes which allow passage of mostly protein-free fluid). In summary, there are four methods for restraining fine particles. In summary, there are five ways to restrain powdered sorbent particles in order to perfuse them with fluid for effective depuration and regeneration: Nanofiber felt bed, Solid extruded block, Sorbent suspension passing through a flat-plate dialyzer, Bidirectional filtration across hollow fiber membranes into a sorbent suspension, and a filtration bed applied by hydraulic flow around a cylindrical filter. Of these five approaches, we have invented the last three.

[0355] While this disclosure has been described as having an exemplary design, the present disclosure may be further modified within the spirit and scope of this disclosure. This application is therefore intended to cover any variations, uses, or adaptations of the disclosure using its general principles. Further, this application is intended to cover such departures from the present disclosure as come within known or customary practice in the art to which this disclosure pertains.