METHOD OF MANUFACTURING AN EMBOLIZING AGENT PRECURSOR
20230233681 · 2023-07-27
Inventors
Cpc classification
A61K49/223
HUMAN NECESSITIES
A61L2300/418
HUMAN NECESSITIES
A61L24/001
HUMAN NECESSITIES
A61L2300/62
HUMAN NECESSITIES
A61K9/0009
HUMAN NECESSITIES
A61K9/1075
HUMAN NECESSITIES
A61K9/16
HUMAN NECESSITIES
A61K41/0028
HUMAN NECESSITIES
A61K31/517
HUMAN NECESSITIES
A61P7/04
HUMAN NECESSITIES
A61K9/19
HUMAN NECESSITIES
A61K31/5377
HUMAN NECESSITIES
A61K31/704
HUMAN NECESSITIES
A61K31/7068
HUMAN NECESSITIES
A61L2300/45
HUMAN NECESSITIES
A61K31/506
HUMAN NECESSITIES
A61L2430/36
HUMAN NECESSITIES
A61K31/475
HUMAN NECESSITIES
A61K41/0047
HUMAN NECESSITIES
A61K31/454
HUMAN NECESSITIES
International classification
A61K9/16
HUMAN NECESSITIES
A61K41/00
HUMAN NECESSITIES
A61K9/19
HUMAN NECESSITIES
Abstract
Disclosed herein are methods relating to manufacturing an embolizing agent precursor. Manufacture of the embolizing agent precursor may involve mixing a first component contained within a first container with a second component contained within a second container, the first component including a plurality of negatively charged gaseous components and a first stabilizer, the second component comprising a plurality of positively charged oil components, a second stabilizer, and a cationic surfactant. Further steps may include mixing the first component with the second component such that the first and second component are held together as a single agglomerated entity.
Claims
1. A method of manufacturing an embolizing agent precursor, comprising: combining a first component contained within a first container with a second component contained within a second container, the first component comprising a plurality of negatively charged gaseous components and a first stabilizer, the second component comprising a plurality of positively charged oil components, a second stabilizer, and a cationic surfactant; mixing the first component with the second component such that the first and second component are held together as a single agglomerated entity; adjusting the size of individual oil components from 1 .Math.m to 5 .Math.ms; modifying the pH of the second component to 6.1-7.4; adjusting the volume concentration of individual oil components to a target concentration of 1-10 .Math.l microdroplets/ml.
2. The method of claim 1, further comprising removing excess stabilizer.
3. The method of claim 1, wherein the first component is produced in-situ in a colloid mill, simultaneously feeding first stabilizer and gas.
4. The method of claim 3, further comprising removing individual gaseous components that have a diameter less than 1 .Math.m.
5. The method of claim 1, further comprising providing the first component in a lyophilised form.
6. The method of claim 1, wherein the second component is prepared from a thermally sterilised lipid dispersion and a sterile filtered oil component.
7. The method of claim 6, wherein the second component is produced in-situ in a colloid mill, simultaneously fed with the lipid dispersion and the oil component, thereby forming a raw emulsion.
8. The method of claim 7, wherein the raw emulsion is size fractionated in an in-line centrifuge, removing excess stabiliser that could contaminate the attraction between first component and the second component.
9. The method of claim 1, wherein the agglomerated entity is stable for at least 1 hour.
Description
BRIEF DESCRIPTION OF THE FIGURES
[0182]
[0183]
[0184]
[0185]
[0186]
[0187]
[0188]
[0189]
[0190]
[0191]
[0192]
[0193]
[0194]
[0195]
[0196]
[0197]
[0198]
[0199]
[0200]
[0201]
[0202]
[0203]
[0204]
[0205]
[0206]
EXAMPLE 1 (E1) - ANALYTICAL TOOLS AND BASIC CHARACTERISTICS OF THE INVENTION
E1-1 Introduction
[0207] The microbubble/microdroplet clusters formed upon combining C1 and C2, i.e. present in DP, are crucial to the critical quality attributes of the composition, i.e. its functionality for delivery of drugs. Hence, analytical methodology to characterize and control the clusters formed with regards to concentration and size, is an imperative tool to assess the current invention as well as for medicinal Quality Control (QC). We have identified three different analytical tools that can be applied for this purpose; Coulter counting, Flow Particle Image Analysis (FPIA) and Microscopy/Image analysis. In the following text, these three analytical methodologies and suitable responses are briefly explained, some basic characteristics of C1, C2 and DP are exemplified, as are some aspects for controlled engineering of these characteristics.
[0208] In addition to these techniques, applied for characterisation of the clusters in the cluster composition, analytical methodology has been developed to study the activation of the clusters in vitro, i.e. the generation of large, activated bubbles upon US irradiation. This meteorology; “Sonometry” is detailed in E1-6. Primary report responses from the Sonometry analysis are number and volume of activated bubbles and their size distribution, both vs. time after activation. Activation responses may also be explored by Microscopy/Image analysis as detailed in E1-5.
E1-2 Components and Compositions Investigated
[0209] The 1.sup.st component (C1) in all the compositions investigated in all examples with exception of E5-2, consisted of per-fluorobutane (PFB) microbubbles stabilised by a hydrogenated egg phosphatidyl serine-sodium (HEPS-Na) membrane and embedded in lyophilized sucrose. HEPS-Na carries a negatively charged head group with an ensuing negative surface charge of the microbubbles. Each vial of C1 contains approximately 16 pL or 2.Math.10.sup.9 microbubbles, with a mean diameter of approximately 2.0 .Math.m.
[0210] The 2.sup.nd component (C2) in the all the compositions investigated in this example consisted of perfluoromethyl-cyclopentane (pFMCP) microdroplets stabilised by a 1,2Distrearoyl-sn-glycerol-3-phosphocholine (DSPC) membrane with 3% mol/mol stearlyamine (SA) added to provide a positive surface charge. The microdroplets in the C2 were dispersed in 5 mM TRIS buffer. The standard formulation of C2 investigated in these studies contains approximately 4 .Math.L or 0.8.Math.10.sup.9 microdroplets per mL, with a mean diameter of approximately 1.8 .Math.m.
[0211] In some cases, to elucidate effects on cluster characteristics, a variety of formulation variables such as SA content, microdroplet size, microdroplet concentration, TRIS concentration and pH was varied in a controlled manner. In case such samples have been used, these aspects are detailed in the text.
[0212] The cluster composition (DP) was prepared aseptically by reconstituting a vial of C1 with 2 mL of C2 followed by 30 s manual homogenisation. 2 mL was withdrawn from a vial of C2 using a sterile, single use syringe and needle. The content of the syringe was added through the stopper of a vial of C1 and the resulting DP was homogenised.
[0213] C1 and C2 manufactured as detailed in Example 9.
[0214] In some cases, in order to compare effects of the cluster composition of the current invention to regular contrast microbubbles, C1 was prepared with pure water instead of C2 to produce an aqueous dispersion of HEPS/PFB microbubbles.
E1-3 Coulter Counting
[0215] Coulter counting is one of the most widely used analytical technique for quantification and size characterization of particulate substances larger than 1 .Math.m and has been shown suitable for QC of medicinal drug products [Sontum, PC. and Christiansen, C., J. Pharm. Biomed. Anal. Vol. 12, No.10, 1233-1241 (1994)]. In brief, a small aliquot of the analyte (e.g. C1, C2 or DP) was diluted/dispersed in a particle free aqueous electrolyte (typically phosphate buffered saline, PBS) and homogenized by continuous stirring. A portion of the diluted sample was then drawn through an aperture in the instrument, over which the resistivity is continuously measured. Each particle that is drawn through the aperture will cause the resistivity to change proportionally to the volume of the particle. During the course of the analysis, the instrument draws a known volume of electrolyte through the aperture, measures and counts each resistivity pulse, and presents the results as number concentration of particles measured vs. size. For the reported analyses a Coulter Multisizer III or IV (Beckman Coulter Ltd.) set up with a 50 .Math.m aperture (measuring range 1 to 30 .Math.m) was utilized. A suitable sample volume was diluted in Isoton II (PBS electrolyte, Beckman Coulter Ltd.) and homogenized by continues stirring throughout the analysis.
[0216] Coulter counting is suitable for quantification of microbubble and microdroplet concentration and size distribution in C1 and C2, and for characterization of particles in DP. As two or more microbubbles/microdroplets in a cluster are counted as a single particle, the formation of clusters upon combining the two components will lead to 1) a reduction in the total number of particles in the system and 2) a shift towards larger sizes. These effects are exemplified in
[0217] A particularly useful response from these measurements is the Reactivity (R) of the cluster composition defined as;
[0218] Where C.sub.C1,C.sub.C2 and C.sub.DP are the number concentration observed in C1, C2 and DP, respectively (in C1, then after reconstitution in 2 mL of pure water). This Reactivity is hence a measure of how many of the individual microbubbles and microdroplets in C1 and C2 that are contained in cluster form in the DP. The Reactivity is also correlated to how large these clusters are (i.e. how many individual microbubbles and microdroplets comprises a single cluster), see E2-5 for further details. E.g. if there are no clustering then C.sub.DP = C.sub.C1 + C.sub.C2 and R = 0% and if all the microbubbles and microdroplets in a cluster composition form a single, large cluster then C.sub.DP ~ 0 and R ~ 100%. From Coulter analysis of C1 (after reconstitution in 2 mL of water), C2 and DP, R can easily be calculated.
[0219] Even though the Coulter analysis is suitable for characterization of the total particle concentration and size distribution in DP, it does not, per se, discriminate between microbubbles, microdroplets or clusters; all entities are counted and sized as “a particle”. In order to differentiate and characterize the clusters specifically, microscopy techniques are necessary.
E1-4 Flow Particle Image Analysis
[0220] Flow Particle Image Analysis (FPIA) is a fully automated microscopy and image analysis technique [Sontum, PC. and Martinsen, E., Abstracts of Eur. Conf. Drug Deliv. Pharm. Tech., Seville (Spain), pp 47, # 25 (2004)]. In brief, a small aliquot of the analyte (e.g. C1, C2 or DP) was diluted/dispersed in a particle free aqueous diluent (water or PBS) and homogenized by continuous stirring. A known portion of the diluted sample was then drawn through a measuring cell in the instrument where a fixed set of micrographs are taken by a CCD camera with a stroboscopic light source. The particles in each frame are automatically isolated and analyzed by the image analysis software, and a variety of morphological parameters are calculated for each particle. In addition, the particle concentration was reported. Of particular interest to the current invention are parameters that discriminate between free microbubbles or microdroplets and clusters of the same. For this purpose the particle size, described as circular equivalent diameter, and their circularity has been used as standard responses. Circular equivalent diameter is defined as the diameter of a circle with an equivalent area as the particle detected. The term “circularity” (C) has its conventional meaning in the field of image analysis and is defined on page 12.
[0221] In addition to numerical responses, the instrument provides a representative selection of micrographs for different size classes; < 5 .Math.m, 5 to 10 .Math.m, 10 to 20 .Math.m and 20 to 40 .Math.m. For the reported analyses a Sysmex 2100 instrument (Malvern Instruments Ltd.) set up with a High Power Field (20X) and measuring range 0.7 to 40 .Math.m was utilized. A suitable sample volume was diluted in water and homogenized by continues stirring throughout the analysis.
[0222] A representative selection of micrographs of individual clusters in the size class between 5 and 10 .Math.m, from analysis of a DP sample made with C2 containing 3.5% stearlyamine is shown in
TABLE-US-00001 Reactivity (R) and concentration (millions/mL) of microbubble/microdroplet clusters in various size classes in the cluster composition at three levels of electrostatic attraction from variable amounts of stearlyamine (SA%) in the stabilizing membrane of the microdroplets S A% R (%) <5 .Math.m 5-10 .Math.m 10-20 .Math.m 20-40 .Math.m 1.5 4 6.5 0.0 0.0 0.0 3.5 21 112.6 3.8 0.1 0.0 5.5 50 84.7 14.0 2.3 0.1
E1- 5 Microscopy / Image Analysis
[0223] As an alternative to the FPIA analysis a more manual microscopy technique coupled with an image analysis software may be employed. For this purpose, a Malvern Morphology G3 system (Malvern Instruments Ltd.) with a 20X objective and a measuring range of 1.8 to 100 .Math.m was utilized. In some cases a 50X objective with a measuring range of 0.5 to 40 .Math.m was utilized. In brief, a small aliquot of the analyte (e.g. C1, C2 or DP) was diluted/dispersed in a particle free aqueous diluent (e.g. water or PBS) and homogenized. The diluted sample was then introduced into a microscopy channel slide (IBIDI .Math.-slide, IBIDI GmBh), with a known channel height of 400 .Math.m and placed under the microscope. The instrument automatically scans a preset area of the slide and a fixed set of micrographs are taken by a CCD camera. The particles in each frame are automatically isolated and analyzed by the image analysis software, and a variety of morphological parameters are calculated for each particle. The total number of particles are reported and from the known scan area and known channel height, the concentration of particles in the analyte can be calculated. As for the FPIA analysis, the circular equivalent diameter and particle circularity was reported. Micrographs of all particles detected can be displayed and evaluated by manual, visual inspection. Hence all clusters can be isolated from e.g. free microbubbles and a full cluster size and circularity distribution can be constructed for the clusters in each sample.
[0224] This methodology can also be used to characterize the activated bubble population, i.e. the cluster composition after ultrasound activation. For this purpose the microscopy slide was immersed in 37° C. water and insonated for 10 s with an ATL 3-2 transducer (center frequency of 2.25 MHz) at a nominal MI of 0.8. Immediately after activation, the slide was placed under the microscope and the analysis was repeated.
[0225] Typical examples of output from this analysis are shown in
E1-6 Sonometry
[0226] In order to study and demonstrate the characteristics of large phase shift bubbles produced after activation of the microbubble/microdroplet clusters present in the composition, an analytical methodology that allows for determination of activated bubble concentration, size and dynamics after activation in a relevant in-vitro system has been developed. The text below describes in detail a method for sizing the activated bubbles in vitro, which produces a measurement of the activated bubble concentration and size distribution from 4 to 80 microns in diameter over time. Measurements are performed every 15 seconds for a period that cover the time of activated bubble growth and dissolution.
[0227] An acoustic transmission technique was used to measure the size distribution dynamics of the activated, large bubble population in-vitro. The acoustic technique requires the measurement of attenuation over a range of frequencies, which are an order of magnitude lower (around 0.2 MHz) than those used for conventional imaging (1-10 MHz). The subsequent conversion to activated bubble size information is based on bubble resonance theory and the solution of the resulting Fredholm integral equation of the first kind, using standard techniques. The associated velocity dispersion data are used to provide a quantitative quality metric with which to assess the performance of the inversion procedure. The technique is based on methods described in the sonar literature to size bubble populations in the upper ocean, with inessential modifications to suit the problem at hand.
[0228] In order to obtain information regarding activated bubble size, the principle acoustic properties are measured as a function of frequency. This data is then inverted to provide size information. The inversion requires an accurate model of the interaction of activated bubbles with the incident sound field. A number of models for the propagation of nonlinear pressure waves in bubbly liquids are available in the literature. Here we restrict measurements to the propagation of low amplitude acoustic waves, which effectively places measurements in the linear region, hence a linear model is employed. We will also restrict consideration and measurements to bubble densities for which the Foldy approximation [Phys. Rev. B, vol. 67, pp. 107-119, 1945] is applicable. Relevant theory is presented in [J. Acoust. Soc. Am 85, pp. 732-746, 1989].
[0229] A low frequency (Panametrics Videoscan SN:267202 part #V1012, 0.25 MHz centre frequency) broadband pulse is directed through a sample cell, reflected from a steel plate (approximately 25 cm from the low frequency transducer), propagates back through the sample and is received by the same transducer. Thus the pulse passes through the sample cell twice. The internal dimensions of the sample cell are: width 7.4 cm, thickness 3.1 cm, height 10.3 cm, giving a total volume capacity of 236.28 cm3. The cell is closed and contains no headspace so that it may be kept at a controlled gas saturation. The temperature to perform the measurements is chosen to be 37° C. to mimic body temperature. The gas saturation in the blood in-vivo is approximately 98 kPa in arterial blood and 90 kPa in venous blood. Coupled with systemic overpressure (~ 100 mmHg) this provides a gas saturation environment of approximately 85% in-vivo. The gas saturation of the sample cell was controlled at 85% to mimic the in-vivo environment. Gentle stirring is incorporated to ensure adequate mixing. Mylar membranes are used to provide acoustically transparent windows. The low frequency source does not activate the clusters. Activation is provided by the high frequency transducer. The bandwidth of the low frequency pulse is able to cover a activated bubble size range from 4 to 80 .Math.m in diameter.
[0230] The inversion procedure is ill-posed in the sense of Hadamard and therefore requires optimisation of the data signal-to-noise ratio. Hence it is appropriate to average as much as is practically possible. 200 consecutive rf A-line signals are recorded at 10 MHz sampling frequency to a nominal 8 bits and comprise one measurement data set. The pulse repetition frequency of the transmission transducer is set to 200 Hz, and thus one second is required for data capture. Data sets are recorded once every 15 seconds and downloaded to a PC for subsequent numerical inversion. 45 such measurement data sets comprise one run, spanning 11 minutes in total.
[0231] Inverting the measured primary acoustic properties to yield activated bubble concentration and size distribution information is based on a simple finite element solution as proposed by Commander and McDonald [J Acoust. Soc. Am. 89 pp. 592-597, 1991]. Details of the inversion algorithm used can be found in [“Solving least squares problems”, Prentice Hall, Chapter 23, p. 161, 1974].
[0232] From the acoustic measurements, acoustic attenuation and velocity as a function of frequency may both be calculated. The velocity data may be regarded as independent to the attenuation data. Only attenuation data is used to calculate the activated bubble size distribution, the velocity data can be used as the basis of an independent check of the estimated activated bubble size distribution. The velocity of a bubbly liquid is highly dispersive around the resonance frequency. This phenomenon may be used to derive a ‘quality’ metric in order to quantitatively infer the accuracy or confidence of the estimated activated bubble size distribution, after [IEEE J. of Oceanic Engineering, vol. 23, no. 3, 1998].
[0233] Primary reports from the meteorology detailed above are activated bubble number and volume concentration, and number and volume weighted bubble diameter, both vs. time after activation.
[0234] Applying the Sonometry analysis, a sample of the cluster composition detailed in E1-2 was analysed.
[0235] In following examples, primary responses evaluated from this analysis are peak activated bubble volume per microdroplet volume or per volume of DP, and volume weighted mean diameter at peak activated volume.
E1-7 Stability of Clusters in the Cluster Composition
[0236] The clusters in the DP are formed and kept by the electrostatic attraction between the microbubbles and the microdroplets. These forces are finite and the clusters may break up after formation through various routes/influences such as mechanical stress or thermal (Brownian) motion.
[0237] For precise and accurate characterization, it is important that the clusters remain stable during the time of analysis. This stability has been investigated with all the methodologies described above. To evaluate stability, 3 to 5 analyses where repeated on a single DP sample covering a timespan of > 5 minutes. No significant change in neither concentration nor size has been observed cross these replicates, proving that the microbubbles, microdroplets and clusters are stable for > 5 minutes under the analytical conditions stated, i.e. after dilution in PBS or water and under continuous homogenization (stirring).
[0238] For use as a medicinal drug product it is imperative that the vital characteristics of the product are kept for a time that enables use. The stability of DP after preparation has been studies with various techniques including FPIA and Sonometry.
E1-9 Formulation Aspects
[0239] A number of different formulation aspects can be explored for controlling the cluster content and size in the DP and for targeting optimal properties. Parameters that can be used to engineer cluster content and size distribution include, but are not limited to; the difference in surface charge between the microbubbles and the microdroplets (e.g. SA% as shown in E1-3): the microdroplet size of C2: the pH: the concentration of TRIS in C2: and the concentration of microbubbles and microdroplets. In addition, chemical degradation of the components, e.g. during prolonged storage at high temperatures, may influence the ability of C1 and C2 to form clusters during preparation of the DP. A brief description of these aspects are given in the following.
[0240] Microdroplet size - Samples of C2 with variable microdroplet size was made from a single batch of raw emulsion by centrifugation and control removal of supernatant and/or sediment after different centrifugation times. After size adjustment, the concentration of all samples was adjusted to the same volume concentration of microdroplets (approx. 4 .Math.l microdroplets/mL). C2 samples with microdroplet size as volume median diameters of 1.8 .Math.m, 2.4 .Math.m and 3.1 .Math.m were prepare and used for preparation of DP with vials from a single batch of C1 and the Reactivity was measured by Coulter counting. The Reactivity was found to increase with decreasing microdroplet size, from 27% at 3.1 .Math.m, to 49% at 2.4 .Math.m and 78% at 1.8 .Math.m. Decreasing microdroplet size hence increase formation of clusters upon mixing of C1 and C2.
[0241] pH - two vials of DP were prepared to a pH of 6.6 and two vials were prepared to a pH of 6.1. The resulting Reactivity, as measured by Coulter counting, was 30-32% and 51-52% for the pH 6.6. and pH 6.1 samples, respectively. A decrease in pH hence increase formation of clusters.
[0242] TRIS concentration - for three samples of C2 from the same batch, the concentration of TRIS was varied from 1 mM to 10 mM. Using a single batch of C1, each sample was used to prepare DP and the concentration of clusters between 5 to 10 .Math.m was measured by FPIA analysis on all samples. The formation of clusters was found to decrease with increasing TRIS concentration, with a lowering of cluster concentration from 6.7 to 3.7 millions/mL, going from 1 to 10 mM TRIS in C2.
[0243] Microdroplet concentration - The formation of clusters upon combining C1 and C2 is also a function of the concentration of microbubbles and microdroplets in the two components, i.e. the ratio of microbubbles to microdroplets. From an intuitive perspective, it seems likely that in a system where the total surface charge presented by the two components balance completely, the result would be that all microbubbles and all microdroplets would form a few, very large clusters (i.e. resulting in a total collapse of the system). We have found that in order to generate a controlled and targeted clustering where most all of the microdroplets are contained in cluster form, and were the clusters formed are of an acceptable size, the total charge presented by the microbubbles should be in excess of the total charge presented by the microdroplets. However, the microdroplet/microbubble ratio must also be above a certain threshold in order to form a significant amount of clusters. Results in Table 2 shows the effect of microdroplet concentration in C2, when used to prepare DP with a fixed concentration of microbubbles in C1 (8 .Math.l microdroplets/mL). As can be noted, we find a strong increase in clustering in terms of Reactivity, and a strong increase in mean cluster diameter, with increasing microdroplet concentration added to a fixed amount of microbubbles.
TABLE-US-00002 Reactivity (R) from Coulter analysis and mean cluster diameter and from microscopy/image analysis in cluster compositions made with C2 with variable microdroplet concentration (C) C (.Math.l/ML) R (%) Mean cluster Diameter (.Math.m) 1.0 8 4.9 1.5 26 NA 3.0 46 5.8 6.0 74 NA 9.0 93 8.5
[0244] Thermal degradation - the SA component in the phospholipid membrane stabilizing the microdroplets in C2 is prone to thermal degradation upon prolonged storage at elevated temperatures, losing its positive charge in the degradation process. The ability to form microbubble/ microdroplet clusters when combined with C1 is hence reduced. Two samples of C1 was place under controlled storage conditions at 4 and 30° C. for three months and used for preparation of DP on which the content of clusters between 5 to 10 .Math.m was measured by FPIA analysis. The cluster contents observed was 29.6 and 0.2 millions/mL for the 4 and 30° C. stored samples, respectively.
E1-10 Size of Activated Bubbles
[0245] The size of the activated bubbles may be of importance to the biological attributes of the administered composition, e.g. safety and efficacy aspects. Whereas naturally depended upon the size of the microdroplets in C2, the activated bubble size is also strongly related to the cluster size.
EXAMPLE 2 (E2) - IN-VIVO STUDIES ON CLUSTER ATTRIBUTES VS PRODUCT EFFICACY
E2-1 Introduction
[0246] Having shown in E1 how to measure important characteristics of the current invention; i.e. for the clusters in the cluster composition, and also how to manipulate and control these, the current example explore which cluster characteristics should be targeted for optimal in-vivo efficacy. In order to reach this objective, two extensive in-vivo studies (Study A and Study B) were performed where the US contrast enhancement obtained after administration of a number of DP samples with different characteristics, was measured in an open chest dog myocardium model. The myocardial enhancement of the US signal was observed after i.v. injection and activation of the composition in the left ventricle. After activation the large phase shift bubbles are trapped in the myocardium capillary network and the US contrast enhancement is a direct measure of the amount of activated bubbles deposited, and hence a measure of the efficacy of the administered sample.
E2-2 Components and Compositions Investigated
[0247] The 1.sup.st component (C1) in the compositions investigated in this example is described in E1-2.
[0248] The 2.sup.nd component (C2) in the all the compositions investigated in this example consisted of perfluoromethyl-cyclopentane (pFMCP) microdroplets stabilised by a 1,2Distrearoyl-sn-glycerol-3-phosphocholine (DSPC) membrane with stearlyamine (SA) added to provide a positive surface charge. The microdroplets in the C2 were dispersed TRIS buffer.
[0249] In order to obtain a significant variance in the cluster characteristics of the cluster composition (DP) formulation variables such as SA content, microdroplet size, microdroplet concentration, TRIS concentration and pH was varied in a controlled manner, as described in E1.
[0250] In Study A the microdroplet size, the SA content (% mol/mol) and the pH was varied in a series of 15 samples as detailed in Table 3. For these samples, the microdroplet and TRIS concentrations were kept constant at approx. 4 .Math.l/mL and 5 mM.
TABLE-US-00003 Variance in C2 component characteristics investigated in Study A C2 sample # Microdroplet mean diam. (.Math.m) SA (%) pH 1 1.0 1.5 7.1 2 1.0 3.5 7.1 3 1.0 5.5 7.1 4 1.8 1.5 6.4 5 1.8 3.5 6.4 6 1.8 1.5 7.1 7 1.8 5.5 7.1 8 2.4 3.5 6.4 9 2.4 5.5 6.4 10 2.4 3.5 7.1 11 2.4 5.5 7.1 12 3.1 3.5 6.4 13 3.1 5.5 6.4 14 3.1 3.5 7.1 15 3.1 5.5 7.1
[0251] In Study B the microdroplet and TRIS concentration and the microdroplet diameter was varied in a series of 15 samples. In addition, one sample was thermally degraded by 3 months storage at 40° C. C2 samples investigated are detailed in Table 4. For these samples the pH was kept constant at 6.2 and the SA content was kept constant at 3%.
TABLE-US-00004 Variance in C2 component characteristics investigated in Study B C2 sample # Microdroplet conc. (.Math.l/mL) Microdroplet mean diam TRIS conc. (mM) 1 4.2 2.1 1 2 4.0 2.1 10 3 3.7 2.1 5 4 3.1 2.0 10 5 2.9 2.0 1 6 2.9 2.0 10 7 3.0 2.1 1 8 2.6 2.0 10 9 2.8 2.0 1 10 3.6 2.1 5 11 3.7 2.4 5 12 3.9 2.4 10 13.sup.1 5.8 1.9 5 14 5.8 2.3 5 15 2.8 2.6 10 1 Sample # 13 was stored 3 months at 40° C. before use.
E2-3 In-Vitro Characterization
[0252] All samples detailed in Tables 3 and 4 where used to prepare and characterize DP as detailed in E1. For all samples the content and size of clusters was determined by FPIA analysis and the content and size of activated bubbles was determined by Sonometry. In addition, for samples detailed in Table 3, the Reactivity was measured by Coulter counting.
E2-4 In-Vivo Procedures
[0253] For both studies, the following in-vivo procedures were applied.
Animal Handling:
[0254] The animal (mongrel or mixed breed dog) arrived on the morning of the experiment day. There was no acclimatization. Anesthesia was induced with pentobarbital (12 - 25 mg kg.sup.-1 i.v.) and fentanyl (1.5 - 2.5 .Math.g kg.sup.-1) and an endotracheal tube was inserted. The animal was transferred to the operating table and was put on volume-controlled mechanical room air ventilation (New England mod. 101 Large Animal Ventilator). When required, O.sub.2-enriched air might be given during some time periods, however not in any of the time intervals from 10 minutes before to 11 minutes after test substance injections.
Anesthesia:
[0255] The animal was kept in general anaesthesia by a continuous i.v. infusion of fentanyl (20 .Math.g kg.sup.-1 h.sup.- .sup.1) controlled by a syringe infusion pump (IVAC model P2000), and pentobarbital (10 mg kg.sup.-1 h.sup.-1) by drip line. The rate of anaesthetics administered might be adjusted somewhat from the nominal value to assure a constant depth of anaesthesia. The depth of anaesthesia was monitored by physiological recordings (heart rate, blood pressure) and by general observation of the animal (signs of muscular activity, breathing efforts, reflexes).
Body Temperature:
[0256] The body temperature was kept constant at 38° by a Harvard homeothermic feedback control unit.
Surgery and Instrumentation:
[0257] A Swann-Ganz catheter for pressure measurements was inserted into the pulmonary artery via the femoral vein and a groin incision. A systemic arterial pressure transducer catheter was inserted into the femoral artery by the same incision. A mid-line sternotomy was performed, and the anterior pericardium was removed. The heart was suspended in a pericardial cradle to avoid compression of the atria and veins. A 0.8 mm Venflon™ cannula was inserted in the right cephalic vein proximal to the elbow joint for injections of test substances.
Physiological Monitoring:
[0258] Arterial and pulmonary artery pressure was measured by SensoNor 840 transducers (Sensonor AS, Horten, Norway) connected to custom-made drift-compensated bridge amplifiers (MAX 420, Maxim Integrated Products, Sunnyvale CA). The amplifier outputs are sampled at 500 Hz and fed to a 8-channel 12-bit ADC card (CIO-DAS 08, Computer Boards) for further processing by PC software (Turbo Pascal 5.0, Borland International). Inhaled and exhaled content of O.sub.2 and CO.sub.2 will be continuously monitored (Capnomac Ultima Respiratory Gas Analyzer) but will not be recorded.
[0259] The following variables are calculated, displayed and recorded for each heartbeat: a) Systolic, diastolic and true mean systemic arterial pressure (SAP), b) True mean pulmonary arterial pressure (PAP) and c) Instantaneous heart rate derived from automated (by software) ECG r-wave detection
Imaging:
[0260] A midline, mid-papillary short axis view of the heart was imaged by an ATL HDI-5000 scanner. A P3-2 transducer was used, the scanner was operated in conventional fundamental B-mode with two focal zones, at the highest frame rate and maximum output power (MI ≈ 1.0). A 30 mm soft silicone rubber pad was used between the transducer surface and the epicardium. All material interfaces are covered by water-based acoustic contact gel.
[0261] The depth of the image was adjusted to the smallest value that will include the whole heart. A dynamic range of 50 dB was used. A pair of digital images from end-diastole and end-systole was stored at each specified point in time. The scanner was left continuously running, except brief periods of cine-loop recalls for storing the images. Digital images are transferred to magnetooptical disk after completion of the experimental session. A PAL VHS video recording of the screen was performed to document the procedures. The identity of the animal and all injections (injection number, substance and dose) should be annotated on the screen.
Injection Techniques and Dosing:
[0262] Prior to each injection, a new vial of C1 was reconstituted with 2 mL of C2. The desired dose of DP (200 .Math.l) was withdrawn and diluted to 2.5 mL with 50 mg/mL TRIS-buffered mannitol (10 mM, pH 7.4). The dose administered was equivalent to 10 .Math.l DP/kg b.w, equivalent to nominally 0.04 .Math.l pFMCP microdroplet and 0.08 .Math.l HEPS/PFB microbubbles per kg. b.w. Injections are performed via a Venflon™ cannula equipped with a rubber membrane port. The cannula and port dead space (about 0.1 mL) was flushed with 5 mL of isotonic saline immediately after each injection.
Experimental Procedures:
[0263] Injections of DP are made via the right cephalic vein, and the resulting myocardial contrast effect is quantified at 90 seconds, 3, 5, 7 and 11 minutes. A baseline reading was performed before each injection. At least 20 minutes was allowed between injections to reduce potential carry-over effects.
Data Analysis and Reporting:
[0264] For each of the specified time points, myocardial contrast effect was read from a large region of interest in the anterior myocardium (MathLab software), tabulated against time and illustrated graphically. The contrast effect at 90 seconds was used as the primary measure of the efficacy for each injection. Contrast intensity values was reported in dB and from these values linear enhancement (Greay Scale units, GS) was calculated.
E2-5 Results From Study A
[0265] The results from in-vitro characterization and myocardial enhancement observed for the 15 compositions investigated are detailed in Table 5. Several important correlations that elucidate the nature and characteristics of the system can be extracted from this data.
[0266] Most importantly, we find an optimum in the Reactivity vs. enhancement correlation, as shown in
[0267] Secondly, we find that the size of the clusters formed is also strongly connected to the Reactivity of the system, as shown in
[0268] In combination, the results displayed in
[0269] Whilst not wishing to be hold to theoretical speculations, possible reasons for these effects could be 1) that the larger masses contained in larger clusters prevent or reduce the activation efficacy or 2) that the larger clusters are retained in the pulmonary circulation after an i.v. injection and hence does not reach the left ventricle where the activation is performed.
TABLE-US-00005 Results from in-vitro characterization and in-vivo performance of investigated compositions - Study A (see text) Sampl e Reactivit y (%) Clusters <5 .Math.m (millions/mL ) Clusters 5 to 10 .Math.m (millions/mL ) Clusters 10 to 20 .Math.m (millions/mL ) Clusters 20 to 40 .Math.m (millions/mL ) Linear enh. (GS) 1 7 27.1 0.1 0.1 0.0 14 2 60 99.7 18.9 1.0 0.1 253 3 87 22.6 8.5 4.5 0.6 102 4 21 111.9 0.8 0.0 0.0 91 5 78 69.8 19.1 2.6 0.0 144 6 11 45.9 0.0 0.0 0.0 12 7 84 36.5 9.5 4.2 0.3 147 8 49 66.0 15.7 2.0 0.0 379 9 70 20.5 9.4 4.2 0.2 180 10 21 112.6 3.8 0.1 0.0 165 11 50 84.7 14.0 2.3 0.1 309 12 27 91.3 9.7 0.2 0.0 433 13 59 24.9 12.3 3.3 0.1 313 14 16 28.4 1.3 0.0 0.0 229 15 28 58.1 11.5 0.6 0.0 286
E2-6 Results From Study B
[0270] The results from in-vitro characterization and myocardial enhancement observed for the 15 compositions investigated are detailed in Table 6. Several important correlations that elucidate the nature and characteristics of the system can be extracted from these data.
[0271] Correlations between cluster concentration between 5 to 10 .Math.m and peak activated bubble volume observed in-vitro vs. myocardial enhancement observed in-vivo are shown in
[0272] Examples 20 to 27 given in WO99/53963 also sites data for myocardial enhancement in a model identical to the one described in E2-4 and the procedures applied are identical. In addition, doses in terms of gas and microdroplet volume administered per kg b.w. are comparable between these studies; WO99/53963 sites 0.35 .Math.l gas and 0.04 .Math.l microdroplets/kg. b.w. whereas in the current example effective doses were 0.08 .Math.l gas and 0.026 to 0.059 .Math.l microdroplets/kg b.w. For comparison then, the range of enhancement observed and cited in Examples 20 to 27 in WO99/53963 has been included in
[0273] As can be noted from
TABLE-US-00006 Results from in-vitro characterization and in-vivo performance of investigated compositions - Study B (see text) Sample Clusters <5 .Math.m (millions/mL ) Clusters 5 to 10 .Math.m (millions/mL ) Clusters 10 to 20 .Math.m (millions/mL ) Clusters 20 to 40 .Math.m (millions/mL ) Sonom. vol. (.Math.l/mL ) Linea r enh. (GS) 1 103.5 11.5 0.5 0.0 1383 369 2 124.1 4.8 0.1 0.0 796 277 3 140.6 10.4 0.2 0.1 1256 284 4 83.0 4.3 0.1 0.0 716 234 5 72.0 5.0 0.3 0.0 627 181 6 37.1 0.6 0.1 0.0 412 130 7 84.1 14.2 0.9 0.0 923 435 8 88.6 1.8 0.1 0.0 359 188 9 120.8 6.5 0.1 0.0 460 207 10 131.6 13.2 0.5 0.0 1073 475 11 125.0 26.4 1.4 0.4 1228 453 12 161.2 29.3 1.4 0.0 1443 661 13 15.9 0.2 0.0 0.0 83 33 14 105.7 29.6 6.3 0.1 1740 693 15 208.5 26.3 1.9 0.1 916 489
E2-7 Multivariate Analysis, Target Cluster Size and Circularity Differentiation
[0274] The results for cluster content in the various size classes and in-vivo enhancement, for all data reported in E2-5 and E2-6, allows for a statistical evaluation of the contribution of the various cluster size classes to in-vivo efficacy. For this purpose a multivariate, principle component analysis was performed. The correlation between the content in the various cluster size classes (X) and enhancement (Y) was determined by partial least squares regression (PLSR). The PLSR algorithm discriminates noise to extract and define true correlations. The validation of PLSR models was performed by applying full cross validation (CV). The CV procedure keeps one sample out followed by testing the precision of the model by estimating (predicting) Y for the excluded sample and compare with the measured Y. The procedure was repeated for each sample, and the number of models was hence equal to the number of samples. By comparing all models derived by cross validation, the significance of X variables were determined by evaluating the variation in regression coefficients originating from each model (p=0.05). The final model is developed from all 30 samples.
[0275] Model accuracy and reliability was done by comparing predicted enhancement and measured enhancement and reliable models were verified by classic statistical quality estimates (r, RMSEC, RMSEP). The evaluation of additional statistical parameters as model leverage and sample distance to model concluded that no critical outliers influenced the model solutions. The Unscrambler software v.9.8, Camo ASA, was used for statistical analysis.
[0276] The results from this analysis are shown in
[0277] The medium Reactivity (R = 46%) sample reported in Table 4 represents DP with the cluster attributes that should be targeted.
E2-8 Conclusions
[0278] According to data and discussions detailed in E1 and E2 we have shown that [0279] Formation of microbubble/microdroplet clusters upon combination of the 1.sup.st component and 2.sup.nd component, i.e. in the cluster composition or pharmaceutical composition, is a prerequisite for its intended functionality in-vivo. [0280] Targeted cluster attributes in terms of size is less than 10 .Math.m, and differentiation from free microbubbles/microdroplets can be designated by a circularity < 0.9.
EXAMPLE 3 (E3) - ACTIVATED BUBBLE SIZE IN-VIVO
E3-1 Introduction
[0281] In order to study and demonstrate the characteristics of large phase shift bubbles produced after activation of the microbubble/microdroplet clusters present in the composition, a methodology that allows for in-vivo determination of activated bubble size and dynamics in a relevant animal model has been developed.
E3-2 Components and Compositions Investigated
[0282] The compositions investigated in this study were as detailed in E1-2.
E3-3 Methodology
[0283] Measurement of activated bubble size distribution and yield of activation was performed in a dog model. The study was approved by the local animal welfare committee. A cannula was placed in the aorta to allow blood flow through an extracorporeal measurement chamber that performs the acoustic bubble sizing. Compound was administered by intra venous injection at 10 .Math.l DP/kg. b.w. and activation provided by a clinical ultrasound scanner imaging the cardiac chambers.
[0284] In order to provide consistency data, compound was also administered via a left atrium cannula with acoustic activation in the cannula, thus providing data that can be directly compared to the same administration (activation in the cannula) into the in-vitro bubble sizing system.
[0285] A mathematical model was developed to calculate the volume of activated bubbles liberated from the measurements performed in the extracorporeal measurement chamber. Results of the model were validated by injecting activated bubbles into the left atrium via a cannula, and comparing the result to the same administration in the in-vitro measurement system.
[0286] At least 15 minutes was allowed between each dosing. Injections were performed with an 18 G needle through a rubber membrane port on a forelimb Venflon™ i.v. cannula. Due to the low dose levels, all i.v. injections were given after 1:10 dilution of DP with an aqueous mannitol/ TRIS solution. The injections were given in about 5 seconds, followed by a flush of saline. On some occasions, injections were performed into the left atrium of the heart via a short polyethylene catheter, either with or without prior activation of the drug product by ultrasound ex vivo. Left atrium injections were slow (20 seconds) to simulate the temporal dispersion of the bolus during normal lung passage. Due to the need for diluting the injected sample in order for the ex vivo ultrasound exposure to penetrate into the fluid, the atrial injections were further diluted with isotonic saline to a total volume of 20 mL.
[0287] The body temperature was kept constant at 38 degree C by a Harvard homeothermic feedback control unit (rectal temperature sensor controlling an electrical heating blanket). A Swann-Ganz catheter for pressure measurements and monitoring of cardiac output (Baxter Vigilance Continuous Cardiac Output (CCO) monitor) was inserted into the pulmonary artery via the femoral vein and a groin incision. A systemic arterial pressure transducer catheter was inserted into the femoral artery via the same incision. A 1.4-mm Venflon™ cannula was inserted in the right cephalic vein proximal to the elbow joint, for injection of test substances. A midline sternotomy was performed, and PEEP was applied to the respirator outlet when entering the pleural spaces. The anterior pericardium was removed, and the heart was suspended by suturing the rim of the remaining pericardium to the wound edges. The auricular appendix of the left atrium was cannulated for injections of activated DP, bypassing the pulmonary circulation.
[0288] The animal was fully anticoagulated by a single intravenous injection of Heparin (1000 i. u. / kg body weight) after complete surgical hemostasis was achieved, and before extracorporal circulation was started. The extracorporal shunt and its associated tubing were filled with isotonic saline and all air was evacuated from the system before the connections to the carotid and jugular catheters were established.
[0289] The pressure inside the acoustic measurement chamber was checked at regular intervals by briefly connecting the pulmonary artery pressure transducer to a side port on the chamber, keeping the transducer at the same elevation level as the chamber.
[0290] No significant deviations in flow or pressure in the shunt circulation were observed during the experiments, and no fibrin clot deposits were observed inside the shunt devices after the experiments. Thus, anticoagulation was adequate.
[0291] A mathematical model of the flow system was developed, in order to estimate the peak concentration in the measurement cell, as a function of flow rate into the cell, and the activated bubble half-life, and bolus half-life. The flow rate may then be adjusted, by altering the flow resistance, in order to ensure adequate dose to the measurement cell. In addition, a mathematical model to estimate the concentration of activated bubbles in the arterial blood from the concentration observed in the measuring cell.
E3- 4 Results
[0292]
[0293] Table 7 below shows the volume-weighted mean activated bubble diameters after i.v. injection measured at arterial conditions (normal arterial blood gas saturation, hydrostatic pressure of about 60 mmHg). The mean value of all observations in the table is 21.4 .Math.m.
TABLE-US-00007 Volume weighted, mean activated bubble diameter at arterial conditions Injection # Dog 1 Dog 2 Dog 3 1 19.7 .Math.m 22.2 .Math.m 20.5 .Math.m 2 21.7 .Math.m 22.3 .Math.m 21.8 .Math.m 3 91.3 .Math.m 21.3 .Math.m 21.6 .Math.m
[0294] Table 8 below shows the rates of activated bubble shrinkage at arterial and venous pressure, given as half-life of gas volume fraction decay in the acoustic measurement chamber. The pressures have been calculated from catheter/ transducer measurements of arterial pressure, and assuming a venous (jugular vein) pressure of zero. Note the faster decay at arterial pressure, this is caused by the elevated partial pressure of all gases inside the activated bubbles, giving larger gradients for outward gas diffusion.
TABLE-US-00008 Arterial and venous chamber pressures and half-lives of activated bubbles Arterial or Venous side Dog 1 Dog 2 Dog 3 Pressure in chamber Half-life Pressure in chamber Half-life Pressure in chamber Half-life Arterial 87 mmHg 21 s 87 mmHg 18 s 88 mmHg 21 s Venous 20 mmHg 38 s 20 mmHg 33 s 20 mmHg 36 s
[0295] The activated bubbles in arterial blood have diameters of 20 - 22 microns, well within the predicted range. After injection of the substance into the left atrium and activating in the left ventricle the activated bubbles become slightly larger, 22 - 25 micron in diameter. Verification of correct measurements and calculations in all animals has been obtained by parallel in-vitro analysis with activation of the injected samples by US irradiation.
E3-5 Conclusions
[0296] Example 3 confirm that the composition is activated within the desired MI range and produces bubble growth and dynamics within the desired size range in vivo after intravenous administration.
EXAMPLE 4 (E4) - INTRAVITAL MICROSCOPY ON DEPOSIT NATURE OF ACTIVATED BUBBLES AND MODELING OF RESPONSE TO US IRRADIATION, COMPARED TO REGULAR MICROBUBBLES
E4-1 Introduction
[0297] In order to further study and demonstrate the characteristics of large bubbles produced after activation in-vivo, a study directly observing individual activated phase shift bubbles within the microcirculation via microscopy of rat mesentery was performed. In addition, to describe the significant differences between the large activated bubbles from the current invention and regular US contrast microbubbles e.g. such as Sonazoid, a theoretical modelling of the volume response to US insonation was performed.
E4-2 Components and Compositions Investigated
[0298] The compositions investigated in this study were as detailed in E1-2.
E4-3 Methodology
[0299] Male Wistar rats were used in the study. The composition was administered intravenously at a dose of 1 mL DP/kg b.w. (i.e. 4 .Math.L/kg b.w. microdroplets and 8 .Math.L/kg b.w. microbubbles). General anaesthesia was administered and maintained with i.v. and i.m. pentbarbital sodium. The rats were intubated, and the tail vein or carotid vein was cannulised for administration of the test formulation. Ultrasound was applied to activate the clusters in the mesentery. The abdomen was opened by means of a vertical midline incision, the rats were then placed in the lateral position on a plastic plate incorporating a round window of cover glass, and the exteriorized mesenteries were placed on the cover glass window. The spread mesenteries were perfused with Krebs-Ringer buffer at 37° C. Ultrasound was applied directly onto the exteriorised mesentery under the objective lens of the microscope. An ultrasound scanner (Elegra; Siemens, Seattle, WA) equipped with a linear probe (7.5L40) was used for ultrasound exposure. Output power was set at maximum corresponding to an MI value of 1.2. Sonar gel was applied between ultrasound transducer and chest wall or the mesentery. Images were recorded on S-VHS or DV tape for subsequent review.
[0300] Simulations of the change in volume of the activated bubbles from the current invention and regular HEPS/PFB microbubbles (C1 reconstituted with water) upon insonation was modelled using the nonlinear bubble model developed by Lars Hoff and described in Acoustic Characterisation of Contrast Agents for Medical Ultrasound Imaging, Kluwer Academic Publishers, 2001, Chapter 8. Simulation parameters for activated phase shift bubble: 8 cycles driving pulse with a MI of 0.2 and frequency of 0.5 MHz, in blood, and 30 micron resting diameter. Simulation parameters for HEPS/PFB microbubbles: 8 cycles driving pulse with a MI of 0.2 and frequency of 5 MHz, in blood, and 3 micron resting diameter.
E4-5 Results
No Ultrasound Activation:
[0301] Two animals were used. No large phase shift bubbles were observed in the mesentery microcirculation after the 6 injections performed.
Ultrasound Activation:
[0302] Three animals were used. Large, activated bubbles were observed after all injections. Activated bubbles were only observed after application of ultrasound. The growth phase of the activated phase shift bubbles could be observed in real-time. The nucleus of the activated bubble grew within a few seconds along with micro vessel blood flow obstruction. No expansion of the micro vessels was observed. The activated bubbles gradually shrank and intermittently advanced in the micro vessels. All activated phase shift bubbles were larger than red blood cells and lodged in the micro vessels and transiently blocked blood flow. All activated bubbles were non-spherical and appeared ellipsoidal in shape, forming against a section of the micro vessel.
[0303]
Simulation of Bubble Dynamics When Exposed to Ultrasound Field:
[0304] Results are shown in
E4-5 Conclusions
[0305] Activated phase shift bubbles with a size of approximately 20 .Math.m were observed when ultrasound activation was applied. No activated phase shift bubbles were observed when ultrasound activation was not applied. The activated phase shift bubbles were transiently (5-19 minutes) deposited in the microcirculation but dislodged as their size decreased.
[0306] Simulation of the volume changes of a phase shift bubble during US insonation shows a three orders of magnitude greater response than a HEPS/PFB microbubble, demonstrating the orders of magnitude greater mechanical work exerted on the tissue by the phase shift bubble.
EXAMPLE 5 (E5) - FORMULATION STUDIES
E5-1 Introduction
[0307] As apparent from E1, the inventors have deliberately chosen to focus formulation studies on variance in C2. This in order to study general effects on the ability to form clusters upon preparation of DP, and hence to obtain control with cluster characteristics and elucidate their importance. It is reasonable to assume that the general formulation aspects/effects shown in E1 apply for a wide variety of microbubble/microdroplet formulation systems. In order to show this, six commercially available microbubble formulations have been tested for preparation of the cluster composition and subsequent activation.
[0308] In addition, two important aspects of the invention; the stability of the pharmaceutical preparation in terms of avoiding spontaneous activation (as noted in WO99/53963) and the ability to load the microdroplets with a therapeutic agent are elucidated in the current example.
E5-2 Cluster Compositions From Commercially Available Microbubble Formulations
[0309] In order to show that the concept of the current invention is applicable to a wide variety of microbubble formulations, DP made from C2 as detailed in E1-2 and six commercially available microbubble products as C1, were tested for cluster content by microscopy/image analysis and activated bubble volume and diameter by Sonometry. The microbubble components investigated as C1 are detailed in Table 9 together with vendors, composition of gas core, stabilizing membrane and pharmaceutical form.
TABLE-US-00009 Commercially available microbubble formulations tested as C1 Product (C1) Vendor Gas core Stabilizing membrane Form Sonazoid GE Healthcare PFB HEPS-Na Lyophilized Optison GE Healthcare PFP Human albumin Aqueous dispersion Sonovue Bracco Spa SF.sub.6 DSPC, DPPG-Na, palmitic acid, PEG4000 Lyophilized Definity Lanteus Medical Imaging Inc. PFP DPPA, DPPC, PEG5000-DPPE, hexadecanoic acid Aqueous dispersion Micromarker VisualSonics Inc. PFB, N.sub.2 Phospholipids, polyethylenglycol, fatty acid.sup.1 Lyophilized PolySon L Miltenyi Biotec GmbH Air Inert, organig polymer.sup.1 Aqueous dispersion .sup.1 Exact chemical composition is not disclosed by the manufacturer.
[0310] For lyophilized forms, preparation of cluster compositions was performed as detailed in E1-2, reconstituting the C1 with a volume of C2 as detailed in the package insert of each formulation (2 mL for Sonazoid, 5 mL for Sonovue and 0.7 mL for Micromarker). For Optison and Definity the microbubbles in a product vial was isolated by removal of infranatant after segregation of the microbubbles and the cluster composition was prepared by adding the same volume of C2 to the vial before homogenisation. For Polyson L, 0.5 mL of homogenized C1 was mixed with 0.5 mL of C2.
[0311] Results for cluster content and activated bubble volume and diameter in the various cluster compositions are stated in Table 10. The C1 component detailed in E1-2 has the same formulation and form as the commercial contrast agent Sonazoid, hence it would be expected that these two agents, when used as C1, would generate a cluster compositions with similar characteristics; as confirmed by the results stated in Table 11. Of the other 5 commercial microbubble product investigated all but Definity yield a cluster composition with significant amounts of clusters which, upon US irradiation, were activated and displayed a significant activated bubble volume. Micromarker, Optison, Sonovue and Polyson, although displaying a strong variance in the chemical composition of the gas core and the stabilizing membrane, show characteristics for their respective cluster compositions which are comparable to those prepared with Sonazoid and C1 as detailed in E1-2. Whilst not wishing to be bound by theoretical considerations it is possible that the reason why the Definity microbubbles does not form clusters with the microdroplet in C2 is the use of the PEG-DDPE component in the stabilizing membrane. This component is likely to create a thick, dens layer of water surrounding the microbubble, thus screening the electrostatic attraction to the microdroplets of C2. An additional finding from this study, when using a 50X objective with a measuring range of 0.5 to 40 .Math.m, was the observation of a significant amount of clusters smaller than 3 .Math.m in the cluster compositions made with Sonovue and PolySon L. These microbubble agents contain a significant amount of small microbubbles compared to C1 as detailed in E1-2, Sonazoid or Optison. In the cluster compositions made with Sonovue and PolySon a significant fraction of clusters formed from ~1 .Math.m microbubbles and ~1 .Math.m microdroplets was observed, and these apparently contributed to the activated bubble volume after US irradiation. Clusters in the size range 1-10 .Math.m should hence be regarded as functional under the current invention.
TABLE-US-00010 Cluster content and activated bubble volume for cluster compositions prepared using various commercially available microbubble formulations as C1 Product (C1) Cluster cone. (milions/mL) Activated bubble volume (.Math.L/.Math.L) Activated bubble diameter (.Math.m) Sonazoid 45 293 40 Optison 23 232 48 Sonovue 32 226 50 Definity 0 0 NA Micromarker 41 293 48 PolySon L 23 167 48
[0312] These results reported above demonstrate that the concept of the current invention is applicable to a wide variety of C1 formulations, both with regards to the composition of the gas core and with regards to the composition of the stabilizing membrane.
E5-3 Spontaneous Activation and US Activation
[0313] The basic nature of the formulation is directed towards a destabilisation of the system i.e. the US induced generation of large phase shift bubbles from the combination of microbubbles and microdroplets. This destabilisation must occur in a controlled manner, in-vivo and at the target site (pathology), and spontaneous growth (activation) upon preparation of DP, or immediately after administration (i.e. in the absence of insonation) is detrimental to the functionality of the invention. WO99/53963 only explore co-administration of the two components but notes that, if the components are mixed prior to administration, avoiding such spontaneous activation of the system is likely to require storage at elevated pressure or low temperature after combination of C1 and C2. The inventors has tried to eliminate these obviously cumbersome and limiting needs to provide a formulation that is stable at ambient conditions. As noted inWO99/53963, based on a theoretical evaluation, it is likely that spontaneous activation is a function of the boiling point (b.p.) of the oil phase and its vapour pressure (v.p). However, the authors of this patent does not identified the possibility that the water solubility of the oil phase may be an even more important contributor to spontaneous destabilisation and bubble growth upon combining C1 and C2. To elucidate these relationships and to provide a solution to this problem, a number of microdroplet phase components (fluorocarbon oils), with a wide range of b.p., v.p. and water solubility, was screened and used for manufacture of C2. These samples were then combined with C1 and assessed for content of spontaneously activated and US activated bubbles. Manufacturing and analysis of these samples are described in the following.
[0314] 641 mg distearoylphosphateidylcholine (DSPC) and 73 mg 1,2-distearoyl-3-(trimethylammonio) propane chloride (DSTAP) were weighed into a 250 mL round bottom flask and 50 mL chloroform was added. The sample was heated under hot tap water until a clear solution was obtained. The chloroform was removed by evaporation to dryness on a rotary evaporator at 350 mm Hg and 40° C., followed by further drying at 50 mm Hg in desiccator over night. Thereafter, 143 mL water was added and the flask again placed on a rotary evaporator and the lipids were rehydrated by full rotational speed and 80° C. water bath temperature for 25 minutes. The samples were placed in refrigerator over night. The lipid dispersion was transferred to a suitable vial and stored in refrigerator until use.
[0315] Emulsions were prepared by transferring aliquots of 1 mL of the cold lipid dispersion to 2 mL chromatography vials. To each of seven vials was added 100 .Math.L of the fluorocarbon oils as detailed in Table 11. The chromatography vials were shaken on a CapMix (Espe, GmbH) for 75 seconds. The vials were immediately cooled in ice, pooled and kept cold until use. Coulter counter analysis was performed to determine the volume concentration of the microdroplets and the emulsions were then diluted with water to 10 .Math.l/mL disperse phase.
[0316] C1 (as detailed in E1-2) was reconstituted in 2 mL of water and mixed together with the C2 samples prepared in a 10 mL tube to a ratio of 10:1 and shaken carefully by hand. The mixture was then diluted with 7 mL water. The samples were evaluated for spontaneously activated and US activated bubbles by microscopy in manual version of the methodology described in E1-4. One mL of this solution was transferred to a microscope cell where the temperature was stabilised to 37° C. after 2 minutes. The cell was set up so that US sonication, using an ATL 3-2 transducer with a center frequency of 2.25 MHz, could be applied to the sample. At 200X magnification the entire cell area was scanned and the content of large (> ~ 15 .Math.m), spontaneously activated bubbles was semi quantitatively assessed by visual inspection. For each sample, a score in the range of 0 to 2 was given, where 0 designate “no or very few large phase shift bubbles observed”, 1 designate “medium number of large phase shift bubbles observed” and 2 designate “large number of large phase shift bubbles observed”. The sample was then insonated for 5 s at a nominal MI of 0.8 and the content of large, US activated bubbles were counted and scored in the same manner. The results from this study is detailed in Table 11 together with the physicochemical characteristics of the compounds investigated.
TABLE-US-00011 Compounds investigated and their physicochemical characteristics; boiling point (b.p., °C), vapour pressure (v.p., torr) at 20° C. and water solubility (logarithm of Molar solubility). Compared to results from assessment of the amount of spontaneously and US activated bubbles. Score 0 = no or very limited, Score 1 = medium and Score 2 = high Compound b.p. v.p. Log W.sub.sol Spont. Act. US Act. Methyl-1.1.2.2-tetrafluoroethyl 34 842.4 -1.4 2 0 2.2.3.3.3-Pentafluroropropyl methyl 46 559.6 -2.0 1 0 Perfluorodimethylcyclobutane 45 579 -5.7 0 1 Perfluoromethylcyclopentane 48 522.7 -5.5 0 1 2H.3H-perfluoropentane 53.6 431.8 -3.2 1 0 1.1.2.3.3.3-Hexafluoropropyl 54.5 418.8 -2.5 2 0 Perfluorohexane 59 359.4 -6.7 0 1 1H.1H.2H-perfluoro-1-hexane 59.5 353.1 -4.6 1 1 1H-perfluorohexane 71 238.6 -5.3 1 1 Perfluoroheptane 82.5 161.2 -7.5 0 2
[0317] The data sited in Table 11 surprisingly shows that the level of spontaneous activation is not significantly correlated to b.p. or v.p., but strongly so to the water solubility of the oil component, as was the level US activation.
E5-4 Drug Loading and Co-Solvents
[0318] In one aspect of the invention, a therapeutic compound is added to the microdroplet oil phase for release at targeted site in vivo upon activation. In order to elucidate concepts to achieve such loading a series of formulation studies were performed. These are briefly summarized in the following.
[0319] Based on screening studies of the various components reported in E5-3, with additional responses such as ease of emulsification, stability of emulsions, availability, quality etc., perfluoromethylcyclopentane (pFMCP) was selected as the primary oil component for manufacture of C2, with a distearoylphosphateidylcholine (DSPC) stabilising membrane added stearlyamine (SA) for positive surface charge. As a starting-point for the study on drug loading, a theoretical evaluation of solubility of different solutes (drugs and chromophores for optical imaging) in pFMCP and a range of other oil components was performed. This evaluation was performed using a state-of-the-art software for assessment of solvent-solute compatibility; Hansen solubility parameters, HSPiP v.4 (Steven Abbott TCNF Ltd.). The HSPiP analysis calculates three basic properties relating to compatibility between substances; Polarity, Dispersion and Hydrogen binding and a distance in this three dimensional space between e.g. a solvent and a solute; the Hansen distance (H.sub.d). The closer the solvent and solute are in this space, the (relatively) better the solubility of the solute in the solvent. Hansen theory predicts that a H.sub.d < 8 represents a soluble “solute in solvent” pair, 8 < H.sub.d < 12 represent partial solubility and H.sub.d > 12 represents non-solubility. This analysis was performed for 1) a series of solvents, selected based on b.p. < 65° C., water solubility < 0.1 M and probable biocompatibility (toxicity), with a large span in Hansen parameters and 2) a series of targeted solutes; chemotherapeutic drugs and molecules suitable for optical imaging. Based on the stated solvent selection criteria, preferred solvents were all partially halogenated hydrocarbons. The miscibility between the solvent and the solubility of the solutes in one of the solvents were experimentally determined. The results from this study are stated in Table 12. In addition to the data stated there, it was found that chlorotrifluoropropane (CltFPr) and dicholorodifluoroethane (dCldFEt) was completely miscible in pFMCP, dicholormethane (dClMe) and tricholormethane (tClMe).
TABLE-US-00012 Physicochemical properties of solvents; boiling point (b.p.), vapour pressure and water solubility (logarithm of Molar solubility and target molecules, Hansen distance (Hd) from pFMCP and tCIMe and miscibility/solubility in tCIMe (see text) Solvent/Target b.p. Log W.sub.sol H.sub.d PFMPC H.sub.d tClMe Miscibility (%) and Solubility (mg/mL) in tClMe pFMCP 46 -5.5 - 11.3 ~ 10% CltFPr 51 -2.4 7.1 6.2 100% (Complete) dCIMe 40 -0 8 11.9 47 100% (Complete) dCldFEt 55 -2.3 8.7 4.3 100% (Complete) tClMe 61 -1.2 11.3 - - Nile Red (dve - - 15.9 6.5 ~ 50 DiR (dve molecule).sup.1 - - 20.3 10.4 ~ 50 Irinotecane - - 15.0 6.6 ~ 350 SN38 - - 18.5 0.6 ~0 Paclitaxel - - 16.7 11.4 ~ 350 Docetaxel - - 23.1 16.1 ~20 Doxorubicin - - 21.2 15.5 ~ 1 Hesperadin - - 20.7 12.1 ~7.5 Idealsib - - 17.5 9.5 ~ 10 Gemcitabine - - 23.1 16.1 ~0 Tosacertib - - 24.6 16.1 ~0 ZM44739.sup.2 - - 17.1 8.4 ~ 1 Afatinib - - 18.1 9.5 ~ 100 .sup.1 DiIC.sub.18(7) (1,1′-Dioctadecyl-3,3,3′,3′-Tetramethylindotricarbocyanine Iodide (Life Tech. Ltd) .sup.2 Experimental aurora kinase inhibitor (Selleckchem Ltd.)
[0320] As can be noted from these results, whereas the calculated H.sub.d fits reasonably with the predicted miscibility between solvents, it is not a good predictor of the absolute solubility of the various target molecules in tClMe. This shows that the Hansen analysis is primarily a tool for the relative solubility of a given molecule in various solvent system, and cannot be used to estimate absolute solubility of various compounds in various solvent systems.
[0321] For seven of the substances sited in Table 12, the measured solubility in tClMe was also correlated to literature values for LogP and LogS. Whereas this analysis indicated, as expected, that the solubility in tClMe is a function of the lipophilicity, these characteristics could not predict absolute solubility. Substances with a LogP < 0.9 and a logS > -2.7 did show no or very low solubility, but for substances with a LogP range of 3.2 to 3.9 the solubility varied from 50 to 350 mg/mL with no covariance to LogP, and for substances with a LogS range of -3.7 to -5.2 the solubility varied from 20 to 350 mg/mL with no covariance to LogS.
[0322] These evaluations show that, whereas lipophilic substances are preferred, the compatibility between any specific therapeutic agent and the invention needs to be tested experimentally.
[0323] None of the target molecules in Table 12 displayed any measurable solubility in pFMCP, hence the use of a co-solvent in order to achieve a functional loading capacity is necessary. As the miscibility of d- and t-ClMe in pFMCP is only some 10%, a “solvent ladder” construction, i.e. the use of a third solvent between tClMe and pFMCP in the Hansen space, is indicated. Based on these considerations a 1:1:1 (by volume) mix of pFMCP, ClrFPr and pFMCP was selected for further studies on C2 loaded with therapeutic or optical imaging compounds.
[0324] The solubility of Nile Red (NR), DiR and Paclitaxel (Ptx) was evaluated in the 1:1:1 mixture of pFMCP, ClrFPr and pFMCP and found to > 5 mg/mL, > 10 mg/mL and > 25 mg/mL, respectively. In addition, a 1:1:2 mixture of said three solvents loaded with Ptx was explored. The solubility of Ptx in this solvent mixture was > 50 mg/mL, showing that the loading capacity can be substantially increased by changing the composition of the oil phase. C2 with a 1:1:1 mixture of these components was manufactured as detailed below.
[0325] A lipid dispersion containing was made by weighting out 250 mg of DSPC with 3% mol/mol SA to 50 mL of water in a 100 mL round bottom flask, hydrated for 30 minutes at 80° C. and allowed to cool. X mg substance (X being 5, 10 and 25 mg for NR, DiR and Ptx, respectively) was weighted out and dissolve in 333 .Math.L tClMe (solution A). 333 .Math.l of solution A was diluted with 333 .Math.l CltFPr + 333 .Math.L pFMCP (solution B). 900 .Math.l lipid dispersion was added to a 1.5 mL centrifuge tube. 100 .Math.l of solution B was added to the lipid dispersion in the centrifuge tube. Emulsification was achieved using a ZoneRay® Dental HL-AH G7 Amalgamator at 3200 rpm for 20s. The resulting emulsion was centrifuged for 5 min at 25 g. After centrifugation, the microdroplets formed a defined sediment layer. The supernatant, containing excess lipid vesicles, was carefully removed, an equivalent volume of 5 mM TRIS in water was added and the microdroplets redispersed by manual shaking. A Coulter analysis was performed and based on the detected volume concentration of microdroplets the emulsion was diluted in 5 mM to 3 .Math.l microdroplets/mL.
[0326] The resulting samples of C2 was assessed by microscopy and Coulter analysis.
[0327] Together with C1 as detailed in E1-2, these C2 samples where then used for preparation of DP which was assessed for Reactivity by Coulter counting, clustering by microscopy and activated bubble size and volume by Sonometry. The observed Reactivity for all samples was in the range of 40-70%, microscopy confirmed the presence of clusters, but showed no evidence of spontaneous activation, the activated bubble volume was in the range of 100-200 .Math.l/.Math.l microdroplets and the activated bubble size was in the range of 42-48 .Math.m. These results demonstrate; [0328] That the microdroplet oil phase can comprise a range of solvents in order to obtain an acceptable drug loading capacity. For this, partially halogenated hydrocarbons are particularly useful. [0329] That a significant fraction (e.g. > 60% v/v) of the solvents can have a significantly higher water solubility (e.g. < 1.Math.10.sup.-1 M) than indicated from E5-2 (< 1.Math.10.sup.-5 M). [0330] That these formulations retain the critical attributes of the concept in the formation of clusters in the cluster composition, their ability to be activated upon insonation and the lack of spontaneous activation.
[0331] C2 samples loaded with DiR as described above were used for assessment of delivery in-vivo (see E8). C2 samples loaded with NR as described above was used for assessment of the expression of the loaded substance upon activation (see E6).
EXAMPLE 6 (E6) - EXPRESSION OF LOADED SUBSTANCE UPON ACTIVATION
E6-1 Introduction
[0332] In order to investigate how a molecular substance, loaded into the microdroplets of C2, will be expressed after activation of the cluster composition, a fluorescence microscopy study was performed. A cluster composition where the microdroplets in C2 had been loaded with Nile Red (NR) dye was activated and studied by fluorescence microscopy.
E6-2 Compounds and Procedures
[0333] C2 loaded with 5 mg/mL Nile Red dye, as detailed in E5-4, was used to prepare a cluster composition as detailed in E1-2. The cluster composition was diluted in water, placed in a microscopy well and activated using a Vscan US scanner (GE Healthcare).
[0334] Images of the activated cluster composition were acquired using a Leica TCS SP8 confocal microscope. The objective used was a HCX IRAPO L 25X water immersion objective with a numerical aperture of 0.95. The fluorescent dye was excited at 539 nm by a tunable white light laser. Emission in the range 570 - 670 nm was detected by a hybrid detector (HyD). The laser speed used was 400 Hz and pinhole diameter was set to 1 AU. Transmission images were acquired simultaneously in another detector, which could be overlaid with the fluorescence images. Intersections and 3D images of the sample were acquired by moving the objective nosepiece stepwise in the z-direction.
E6-3 Results
[0335]
[0336] As can be noted from this figure, after activation the loaded NR is presented at the liquid/gas interface. After activation in vivo, if loaded with therapeutic substance, such substance would be in close contact with the endothelial wall and hence accessible for extravasation.
EXAMPLE 7 (E7) - DEPOSITION OF ACTIVATED BUBBLES IN TUMORS
E7-1 Introduction
[0337] In order to further study and demonstrate the characteristics of large bubbles produced after activation in a tumour model, a study imaging activated phase shift bubbles in subcutaneous prostate cancer PC-3 tumour xenografts in a murine model was performed, demonstrating the deposit nature and marked difference in contrast enhancement kinetics from free flowing HEPS/PFB microbubbles.
E7-2 Components and Compositions Investigated
[0338] The compositions investigated in this study were as detailed in E1-2.
E7-3 Methodology
[0339] 16 Female Balb/c nude mice were used. Before tumour implantation, mice were weighted, anesthetized with isoflurane, and ear marked. 100 .Math.L cell suspension containing 3.Math.10.sup.6 PC-3 cells were slowly injected subcutaneously on the lateral side of the left hind leg between the hip and the knee.
[0340] The mice were administered surgical anesthesia by subcutaneous injection of a mix of Fentanyl (0.05 mg/kg), Midazolam (5 mg/kg), and Medetomidine (0.5 mg/kg). An intravenous cannula (BD NeoflonTM 24 GA) was placed in the tail vein. Patency was verified by injection of a slight amount (~20 .Math.L) of 0.9% sodium chloride for injection after which a small amount of (~10 .Math.L) heparin (10 U/mL) was injection to prevent clotting. The hub of the cannula was filled with 0.9% sodium chloride for injection to eliminate any dead space and closed with a cap. The cannula was secured to the tail with surgical tape.
[0341] Three commercial ultrasound imaging systems were used. The tumour was imaged for all experiments with a high frequency small animal imaging system Vevo 2100 (VisualSonic Inc.) with a MS250 transducer (16-18 MHz). The cluster composition was activated in-vivo either with a Vivid E-9 clinical imaging system (GE Healthcare) using a 2 MHz imaging probe with MI setting of 0.28, or a Vscan 1.2 clinical imaging system (GE Healthcare) with a 2 MHz imaging probe with an nominal MI setting of 0.8.
[0342] The 16 animals were split into 4 groups of 4 animals in each group. The activation system and dose for the groups are stated in Table 13.
TABLE-US-00013 Groups investigated with activation procedure and dose Group Activation Dose 1 Vivid E9. MI 0.28 1.5 .Math.L/kg pFMCP + 4 .Math.L/kg HEPS/PFB 2 Vscan. MI 0.8 1.5 .Math.L/kg pFMCP + 4 .Math.L/kg HEPS/PFB 3 Vivid F9. MI 0.28 6 .Math.L/kg pFMCP + 16 .Math.L/kg HEPS/PFB 4 Vscan. MI 0.8 6 .Math.L/kg pFMCP + 16 .Math.L/kg HEPS/PFB
[0343] The prepared mice were place on a handling table (with temperature control set to 37° C.), on its right side. The left leg was lifted horizontally, supported by a piece of cloth and fixated with surgical tape. Ultrasound gel was richly applied unto the tumour and a water-bath-bag was placed on top of the tumour. Imaging was performed with the Visualsonics Vevo 2100 imaging system with the transducer placed in the water bath with the imaging transducer held in a fixed scan plane. Activation of the cluster composition was performed with an additional transducer (either the Vivid E9 or Vscan) for 75 seconds starting from the time of injection of the cluster composition.
[0344] Activation of the cluster composition produces contrast echoes which remain stationary in the ultrasound image for several minutes. The number of stationary contrast signals was counted per unit area of the tumour imaged in the scan plane. Assuming a scan plane thickness of 0.2 mm the number of phase shift echoes per unit volume of tumour was derived.
E7-4 Results
[0345] The tumour is hypo echogenic in the ultrasound image. Typical tumour images are shown in
TABLE-US-00014 Estimated mean number of phase shift stationary echoes per mL tumour tissue, and standard deviation (SD) Group Mean number per SD 1 (MI 0.28. low dose) 3685 2188 2 (MI 0.9. low dose) 6675 1701 1 (MI 0.28. high dose) 10705 6394 4 (MT 0.9.sub.. high dose) 12597 7884
[0346] A two-way analysis of variance was applied for dose and type of activation (Vivid E9 or Vscan). There is a statistically significant difference for dose with p=0.024, and an insignificant difference for activation transducer, p=0.352.
E7-5 Conclusions
[0347] The cluster composition was activated with two different clinical imaging systems, a Vivid E9 with a 2 MHz probe and MI of 0.28, and a Vscan with a 2 MHz probe and MI of 0.8, and tumours imaged with a high frequency (16-18 MHz) small animal ultrasound imaging system. All procedures produced stationary contrast echoes in the tumours. These contrast echoes remain stationary in the ultrasound image for several minutes as opposed to the transient contrast echoes from HEPS/PFB microbubbles. They are not destroyed with burst imaging sequences designed to destroy HEPS/PFB microbubbles. These observations are consistent with phase shift bubble deposition in the tumour tissue. A statistically significant dose response was observed (p=0.24) and the amount of deposition was not statistically different when activated with the different clinical imaging systems with MI of 0.28 and 0.8 (p=0.352).
EXAMPLE 8 (E8) - DELIVERY OF CO-INJECTED OR LOADED SUBSTANCES TO TUMORS
E8-1 Introduction
[0348] In order to demonstrate the ability of the current invention to enhance delivery of molecules in-vivo, studies in a mouse PC-3 xenograph tumour model were performed. Three model systems were explored; co-injection of DP and Evans Blue dye: co-injection of DP and Licor CW800 EPR agent: and injection of DP where a DiR-dye had been loaded into the microdroplet component (C2). Evans blue is a fluorescent dye that binds to albumin protein when injected i.v. Under physiologic conditions, the endothelium is impermeable to albumin, and Evans blue bound albumin remains confined within blood vessels. Thus, Evans blue is often used as a model compound in drug delivery studies [Bohmer et al., J Controlled Release, 148, Issue 1, 2010, pp. 18-24]. Enhanced permeability and retention (EPR) is a common characteristic of tumour vasculature. The vascular endothelium in the tumour microenvironment is often discontinuous, allowing molecules to diffuse into the surrounding tumour tissue. The commercially available (LiCor Biosciences Inc.) IR dye 800CW PEG contrast agent (25-60 kDa) is a non-specific imaging agent intended accumulate in tumours due to the EPR effect. DiR dye is a commercially available (Life Technologies, Thermo Fisher Scientific Inc.) near IR fluorescent, lipophilic carbocyanine DiOCis(7) dye which is weakly fluorescent in aqueous conditions but highly fluorescent and photostable when incorporated into e.g. cell membranes. Thus the standard techniques of extraction and quantification of Evans Blue in tissue, and optical imaging with the 800CW PEG and DiR dyes, were employed as model compounds for in-vivo demonstration of drug delivery with the current invention.
E8-2 Components and Compositions Investigated
[0349] The compositions investigated in this study were as detailed in E1-2 (co-injection models) and E5-4 (DiR loaded).
E8-3 Methodology
[0350] Female Balb/c nude mice were used in the study. Before tumour implantation, mice were weighted, anesthetized with isoflurane, and ear marked. 100-.Math.l cell suspension containing 3.Math.10.sup.6 PC-3 cells were slowly injected subcutaneously on the lateral side of the left hind leg between the hip and the knee.
[0351] The mice were administered surgical anesthesia by subcutaneous injection of a mix of Fentanyl (0.05 mg/kg), Midazolam (5 mg/kg), and Medetomidine (0.5 mg/kg). An intravenous cannula (BD NeoflonTM 24 GA) was placed in the tail vein. Patency was verified by injection of a slight amount (~20 .Math.L) of 0.9% sodium chloride for injection after which a small amount of (~10 .Math.L) heparin (10U/mL) was injection to prevent clotting. The hub of the cannula was filled with 0.9% sodium chloride for injection to eliminate any dead space and closed with a cap. The cannula was secured to the tail with surgical tape.
[0352] The hind limb of the mouse was placed in a water bath with two US transducers poised for insonation of the tumour. Ultrasound activation of the cluster composition was provided by a Vscan with 2 MHz probe and nominal MI of 0.8. Subsequent ultrasound exposure was applied using 500 kHz custom made transducer (Imasonic SAS), 8 cycle pulses with a pulse repetition frequency of 1 kHz at MI ranging from 0.1 to 0.8.
Evans Blue
[0353] 50 .Math.l Evans Blue (50 mg/kg) was injected followed immediately by 50 .Math.L of the cluster composition containing a nominal 1.5 .Math.L pFMCP microdroplets + 4.0 .Math.L HEPS/PFB microbubbles per mL, or 4.0 .Math.L HEPS/PFB microbubbles per mL only. Activation was provided by a Vscan clinical ultrasound scanner with a 2 MHz probe for 45 seconds starting from the injection time. This was subsequently followed by 5 minutes 500 kHz ultrasound irradiation at an MI of 0.1 or 0.2. 30 minutes after treatment the animals were sacrificed, tissue samples; tumour, thigh muscle from the treated leg and thigh muscle from the contra lateral untreated leg, were harvested and Evans Blue content extracted and quantified. Three animals were tested in each group, all with 45 s activation using the VScan probe. Groups and variables are given in table 15.
TABLE-US-00015 Groups investigated with co-injection of Evans Blue dye; US procedure and test items. US activation was performed on all animals, three animals per group Group Subsequent US irradiation Test item 1 MI 0.1, 5 min Cluster composition MI 0.2, 5 min Cluster composition 2 None Cluster composition 3 MI 0.2, 5 min HEPS/PFB microbubbles
LiCor CW800 EPR Agent
[0354] LiCor CW800 EPR agent was administered at a dose of 5 nmol/kg body weight followed immediately by 50 .Math.L of the cluster composition containing a nominal 1.5 .Math.L pFMCP microdroplets + 4.0 .Math.L HEPS/PFB microbubbles per mL. Activation was provided by a Vscan clinical ultrasound scanner with a 2 MHz probe for 45 seconds starting from the injection time. This was subsequently followed by 5 minutes 500 kHz ultrasound irradiation at an MI of 0.2. Whole body epifluorescence imaging was performed with a Pearl Impulse imaging system up to 12 hours post administration. Animal groups and numbers are given in Table 16.
TABLE-US-00016 Groups investigated with co-injection of LiCor CW800; US procedures. All animals were dosed with cluster composition Group # animals Activation Subsequent US Irradiation 1 3 None None 2 4 45 seconds None 3 5 45 seconds MI 0.2, 5 min
[0355] A region of interest was drawn over the tumour in the epifluorescence image and the mean intensity calculated. A commensurate region of interest was also drawn over the untreated, contralateral thigh in approximately the same location on the leg. A dimensionless ratio was calculated of the average image intensity in the tumour region area divided by the average image intensity in the untreated leg. The area under the curve of this ratio was calculated and integrated from the 1 minute to 1 hour time points.
DiR Loaded Cluster Composition
[0356] 50 .Math.L of the cluster composition containing a nominal 1.5 .Math.L pFMCP microdroplets loaded with 10 mg/mL DiR dye + 4.0 .Math.L HEPS/PFB microbubbles per mL was administered. Activation was provided by a Vscan clinical ultrasound scanner with a 2 MHz probe for 45 seconds starting from the injection time. This was subsequently followed by 5 minutes 500 kHz US irradiation at an MI of 0.2. The ultrasound fields were applied to the tumour bearing left legs of the animals. The control group received the same DiR loaded cluster composition and handling procedures but no ultrasound exposure. Animal group details are given in Table 17.
TABLE-US-00017 Groups investigated with injection of DiR loaded cluster composition; animal numbers and US procedures. All animals were dosed with cluster composition Group # animals Activation Subsequent US irradiation 1 3 None None 2 4 45 seconds MI 0.2, 5 min
[0357] Epifluorescent images were acquired with the Pearl Impulse fluorescence imaging system both pre-injection and 1 minute post treatment (approximately 7 minutes post-injection) with standardised image acquisitions to allow quantitative comparisons. Regions of interest were drawn over the tumour on the left thigh, and a commensurate region of interest drawn on the non-tumour bearing right thigh of approximately the same size and anatomical location. The mean fluorescence intensity in the regions was recorded. As primary response, the difference in the fluorescence intensity between the pre-injection image and the post-treatment image was assessed.
[0358] A two way analysis of variance was performed with factors of tumour vs non-tumour bearing leg, and US irradiation vs no US irradiation.
E8-4 Result
Evans Blue
[0359] The Evans Blue was extracted and quantified from the tissue samples (mg/mL tissue). The concentration in the treated thigh muscle was divided by the concentration in the untreated thigh muscle for each animal (matched pair) to provide a dimensionless ratio of the increased uptake in the treated muscle. A one way ANOVA was applied to the data and results are given in Table 18. There was a statistically significant, approximate doubling in the Evans Blue uptake in the leg treated with the activated cluster composition with subsequent low frequency applied. For the other groups, no statistically significant increase in uptake was observed.
TABLE-US-00018 Uptake ratios (mean and standard deviation) for treated vs. untreated muscle tissue in different groups Group Mean SD 1 - Cluster composition, activation and subsequent US irradiation 2.0 0.3 2 - Cluster composition and activation only 1.3 0.4 3 - HEPS/PFB microbubbles, activation and US irradiation 1.1 0.2
[0360] Tumour samples were taken from groups 1 and 2. The Evans Blue concentration was divided by the concentration of Evans Blue in the untreated thigh muscle tissue sample to provide a dimensionless ratio describing increase in uptake. A 2 sample t-test was applied with assumed equal sample variance. The results are shown in Table 19. There was an increased uptake in the tumour tissue compared to the untreated thigh muscle of approximately 3.4 to 1 for the tumour with 500 kHz ultrasound applied after activation and approximately 2 to 1 without the application of 500 kHz ultrasound subsequent to activation.
TABLE-US-00019 Uptake ratios (mean, standard deviation and standard error of mean) for treated tumour tissue vs. untreated thigh tissue in different groups Group Mean SD SEM 1 - Cluster composition, activation and subsequent US irradiation 3.4 1.0 0.4 2 - Cluster composition and activation only 2.0 0.4 0.2
Optical Imaging With LiCor CW800 EPR Contrast Agent
[0361] Typical epifluorescence images are shown in
[0362]
[0363] The ratio of the average intensity in the tumour region to the average intensity in the untreated leg was calculated to create a dimensionless Target to Background (TBR) ratio, and the area under the TBR curve was integrated from 1 minute to 1 hour post treatment. Imaging was performed at 1 minute, 30 minutes and 60 minutes time points for all animals. The results are tabulated in Table 20.
TABLE-US-00020 Area under curve (AUC) for TBR uptake ratios (mean and standard deviation) for treated tumour tissue vs. untreated thigh tissue in different groups Group Mean AUC SD 1 - No activation, no subsequent US irradiation 63.7 4.5 2 - Activation only 77.0 7.3 3 - Activation and subsequent US irradiation 94.0 7.5
[0364] An analysis of variance was applied to the three treatment groups with resulting p value of <0.001. Contrasts were applied between groups 1 and 2 with p value of 0.037, and between groups 2 and 3 with p value of 0.005. There is thus a statistically significant (at the 0.05 level) increase in the area under the curve between groups 1 and 2 and between groups 2 and 3.
DiR Loaded Microdroplet Component of the Cluster Composition
[0365] Typical post-treatment epifluorescence images are shown in
TABLE-US-00021 Mean increase ± SD in fluorescence intensity from pre-injection values in different groups Group Left (tumour) leg Right leg 1 - no activation or US irradiation -0.02 ± 0.20 0.17 ± 0.19 2 - activation and subsequent US irradiation 2.83 ± 0.57 0.23 ± 0.31
[0366] A two way analysis of variance gives a p value <0.001 for increased fluorescence intensity in the tumour bearing leg when ultrasound activation and subsequent irradiation was applied. These results demonstrate localised delivery and uptake of the fluorescent dye molecule loaded into the oil phase of the microdroplet component (C2).
E8-5 Conclusions
[0367] There was a statistically significant increase in delivery of Evans Blue dye to muscle and tumour tissue at the 0.05 level upon co-injection with the cluster composition followed by US activation and further US irradiation. An approximate doubling of Evans Blue was observed in the treated muscle tissue compared to no activation and no subsequent US irradiation. No significant increase in delivery of Evans Blue was observed with the administration of HEPS/PFB microbubbles only, and application of the activation and subsequent ultrasound exposure. The uptake in tumour tissue compared to untreated muscle tissue increased by a factor of 2 upon activation only and a factor of 3.4 upon activation and subsequent US irradiation.
[0368] There was a statistically significantly increase in delivery of Licor CW800 EPR agent to tumour tissue upon co-injection with the cluster composition and ultrasound activation compared to no ultrasound activation. There was a further statistically significantly increase in delivery upon subsequent US irradiation with subsequent ultrasound irradiation for enhanced delivery, compared to activation alone.
[0369] A statistically significant increase in fluorescence intensity in the tumour bearing leg was observed when activation and subsequent US irradiation was applied after injection of a cluster composition where DiR dye had been loaded into the oil phase of the microdroplets in C2. This demonstrates localised delivery of a molecular payload from the microdroplets, confirming targeted spatial release and uptake to areas exposed to the ultrasound procedure.
EXAMPLE 9 (E9) - MANUFACTURE OF THE COMPONENTS AND PREPARATION OF THE CLUSTER COMPOSITION
[0370] Both components were manufactured aseptically.
[0371] C1 - A raw dispersion of microbubbles were prepared from a sterile lipid dispersion and sterile gas component. The lipid dispersion was thermally sterilised in a bulk vessel and the gas was sterile filtered. The complete production line was steam sterilised. The microspheres were produced in-situ in a colloid mill, simultaneous fed with lipid dispersion and gas. The intermediate product (raw dispersion) was then size fractionated in a flotation vessel, diluted to target microbubble concentration with an aqueous solution of lyophilisation-protecting agent, filled aseptically and lyophilised.
[0372] C2 - The microdroplet emulsion was prepared from a sterile lipid dispersion and a sterile oil component. The lipid dispersion was thermally sterilised in a bulk vessel and the oil component was sterile filtered. The complete production line was steam sterilised. The microdroplets was produced in-situ in a colloid mill, simultaneously fed with the lipid dispersion and oil component. The raw emulsion was then size fractionated in an in-line centrifuge, diluted to target microdroplet concentration with an aqueous solution of TRIS buffer, and filled aseptically.
[0373] Three consecutive batches of each component was manufactured and subjected to sterility testing according to Ph.Eur and USP. All six batches passed the sterility test.
[0374] DP - The cluster composition was prepared aseptically by reconstituting a vial of C1 with 2 mL of C2 followed by 30 s manual homogenisation. 2 mL was withdrawn from a vial of C2 using a sterile, single use syringe and needle. The content of the syringe was added through the stopper of a vial of C1 and the resulting DP was homogenised.