IOP PUBLISHING

PHYSICS IN MEDICINE AND BIOLOGY

Phys. Med. Biol. 53 (2008) 6605–6618

doi:10.1088/0031-9155/53/22/020

Use of conductive gels for electric field homogenization increases the antitumor efficacy of electroporation therapies Antoni Ivorra1,5,6 , Bassim Al-Sakere2,3,5 , Boris Rubinsky1,4 and Lluis M Mir2,3 1 Department of Bioengineering, Department of Mechanical Engineering and Graduate Program in Biophysics, University of California at Berkeley, Berkeley, CA 94720, USA 2 UMR 8121 CNRS-Institut Gustave-Roussy, Villejuif 94805, France 3 Univ Paris-Sud, UMR 8121, Villejuif 94805, France 4 Center for Bioengineering in the Service of Humanity and Society, School of Computer Science and Engineering, Hebrew University of Jerusalem, Givat Ram, Jerusalem, 91904, Israel

E-mail: [email protected]

Received 21 June 2008, in final form 13 October 2008 Published 31 October 2008 Online at stacks.iop.org/PMB/53/6605 Abstract Electroporation is used in tissue for gene therapy, drug therapy and minimally invasive tissue ablation. The electrical field that develops during the application of the high voltage pulses needs to be precisely controlled. In the region to be treated, it is desirable to generate a homogeneous electric field magnitude between two specific thresholds whereas in other regions the field magnitude should be as low as possible. In the case of irregularly shaped tissue structures, such as bulky tumors, electric field homogeneity is almost impossible to be achieved with current electrode arrangements. We propose the use of conductive gels, matched to the conductivity of the tissues, to fill dead spaces between plate electrodes gripping the tissue so that the electric field distribution becomes less heterogeneous. Here it is shown that this technique indeed improves the antitumor efficacy of electrochemotherapy in sarcomas implanted in mice. Furthermore, we analyze, through finite element method simulations, how relevant the conductivity mismatches are. We found that conductivity mismatching errors are surprisingly well tolerated by the technique. Gels with conductivities ranging from 5 mS cm−1 to 10 mS cm−1 will be a proper solution for most cases. (Some figures in this article are in colour only in the electronic version)

5 6

These authors contributed equally to this work. Author to whom any correspondence should be addressed.

0031-9155/08/226605+14$30.00

© 2008 Institute of Physics and Engineering in Medicine

Printed in the UK

6605

6606

A Ivorra et al

1. Introduction Electroporation, or electropermeabilization, is the phenomenon in which cell membrane permeability to ions and macromolecules is increased by exposing the cell to short (microsecond to millisecond) high electric field pulses. Reversible electroporation of living tissues is the basis for different therapeutic maneuvers on clinical use or under study (Mir 2000) such as the in vivo introduction of genes into cells (electrogenetherapy) (Jaroszeski et al 2000, Dean 2005, Mir et al 2005) and the introduction of anti-cancer drugs into undesirable cells (electrochemotherapy) (Mir et al 2006, Sersa 2006, Marty et al 2006, Miklavcic et al 2006). More recently, irreversible electroporation (IRE) has also found a use in tissues as a minimally invasive surgical procedure to ablate undesirable tissue without the use of adjuvant agents (Davalos et al 2005, Edd et al 2006, Al-Sakere et al 2007, Rubinsky 2007). Electroporation is a dynamic phenomenon that depends on the local transmembrane voltage. It is generally accepted that, for a given pulse duration and shape, a specific transmembrane voltage threshold exists for the manifestation of the electroporation phenomenon (from 0.5 V to 1 V). This leads to the definition of an electric field magnitude threshold for electroporation (Erev); only the cells within areas where E  Erev are electroporated. If a second threshold (Eirrev) is reached or surpassed, electroporation will compromise the viability of the cells because electroporation becomes irreversible (cells do not reseal). A larger threshold can also be defined (Ethermal) for the manifestation of thermal damage caused by the Joule effect. This is particularly relevant in the case of IRE ablation techniques: if irreversibility threshold is surpassed but thermal threshold is not reached then cells are destroyed but tissue scaffold is spared and that facilitates post-treatment healing (Rubinsky 2007). It is quite obvious from the above that precise control over the electric field that develops in tissues is important for electroporation therapies (Gehl et al 1999, Miklavcic et al 1998, 2000, Valic et al 2004). The goal is to generate a rather homogeneous electric field magnitude (Emin  E < Emax) in the region of interest and an electric field magnitude as low as possible in the regions not to be treated. Note that we have indicated that the electric field magnitude in the region of interest must be between Emin and Emax instead of Erev and Eirrev. The Erev and Eirrev labels are usually employed to denote the minimum electric field magnitudes at which reversible electroporation and irreversible electroporation can be manifested, but that does not imply that all the cells will experience reversible or irreversible electroporation. In other words, even if the electric field magnitude at a certain area is slightly larger than Erev, that does not guarantee that all the cells in that area will experience enough reversible electroporation. Multiple factors, such as cell size, cell shape and intercellular spacing, modulate the induced transmembrane voltage. According to known experimental results, it is possible to define an Etarget that represents the optimal electric field magnitude for a specific treatment. Then we also define a tolerance around this value. That is, in the region to be treated, the electric field magnitude must be within this tolerance interval in order to achieve successful treatment. Currently, optimization of the electric field distribution during electroporation is done through design of optimal electrode setups (Hofmann 2000, Edd and Davalos 2007, Zupanic et al 2008). However, there are situations in which an electrode setup alone is not sufficient for obtaining an optimal electrical field, particularly in situations such as the electroporation of irregularly shaped tissues or when the protection of specific tissue regions is required. In Ivorra and Rubinsky (2007a), we introduced the concept of using gels with specific conductivities in order to optimize the electric field distribution in different electroporation scenarios. One of those scenarios was the treatment of bulky superficial tumors. Although electrochemotherapy with plate electrodes is achieving impressive results with the treatment

Conductive gels for field homogenization in electroporation

6607

of such tumors (Mir 2000), some failures could be attributed to heterogeneities in the electric field distribution. A particularly interesting case is that of hard tumors that cannot be squeezed properly between the electrodes so that the geometry is not uniform between the plates and, as a consequence, the electric field distribution is heterogeneous (simulation results reported here help to understand this statement). The solution we propose is extremely simple: to fill all the space between both plate electrodes with a gel whose conductivity is the same as that of tissue. Then the whole material between the plates will become less heterogeneous in electrical terms and the generated electric field distribution will also be less heterogeneous. Conductive gels are already being used in electrochemotherapy in order to improve the electrical contact between the electrodes and the tissues. Nevertheless, this must not be confused with what we are proposing here. Gels for contact improvement should be highly conductive and must be applied in a thin layer. On the other hand, gels for electric field homogenization must have their conductivity matched to that of tissues and should be applied generously, filling all the space between electrodes. We describe here numerical and experimental studies demonstrating the validity and usefulness of this proposal. We also report the electrical conductivity of some commercial gels that were tried to assess their suitability for the proposed technique. While the structure of this report follows a classical outline (Introduction, Methods, Results, Discussion and Conclusions), the Methods and Results sections are divided into three subsections: (1) analysis of the technique based on computer simulations by the finite element method, (2) conductivity measurements of some commercial gels and (3) experimental study of the technique in an animal model (subcutaneous tumor in mice). 2. Methods 2.1. Computer simulations We employed the finite element method (FEM) to compute the electric field distribution under the assumption of static currents and fields. In the areas that result in electric field magnitudes larger than Emin but lower than Emax, we considered that electroporation is successfully achieved. This modeling methodology has been used by previous researchers in the field (Sel et al 2003, Dev et al 2003, Sugibayashi et al 2001) and it has been proven empirically that it is able to predict reasonably well the outcome of electroporation treatments (Edd et al 2006, Miklavcic et al 2000). Here, in order to improve the realism of the models, we consider that the tissue conductivity is not constant but it is electric field dependent. This feature was not taken into account until recently by researchers in the field but it has been shown to have a significant effect on electric field distribution computations (Sel et al 2005). Moreover, in contrast to a previous conference communication (Ivorra and Rubinsky 2007c), here the tissue conductivity is not considered homogeneous; three different tissues with their own conductive properties are considered: skin, tumor and muscle. The key idea of the FEM is the decomposition of an arbitrary geometry into small simple elements in which it is possible to solve the differential equations related to the phenomena under study. Given the appropriate boundary conditions, the solutions are then assembled and an approximate solution for the complete geometry is provided. In our case, the solved equation for each element is Poisson’s equation: (1) −∇ · (σ ∇V − J e ) = 0 e where σ is the conductivity, V is the voltage and J is a vector denoting the externally generated current density.

A Ivorra et al

6608 (a)

(b)

Figure 1. (a) Geometry of the model employed in the FEM simulations; (b) parts of the model. Electrode thickness is 1 mm.

The specific FEM tool used here was COMSOL Multiphysics 3.4 (www.comsol.com) and the mode chosen for the simulations was ‘3D conductive media DC’. The boundary conditions were all insulating on the external surfaces. A mesh of 46 964 tetrahedral elements was automatically generated by the FEM tool. The geometry of the analyzed case is shown in figure 1. Up to a point, it tries to model the electrochemotherapy in vivo experiments reported in the next sections. The electrode separation distance is 4.4 mm and the applied voltage is 572 V. In the ideal parallel plates’ setup, that is, two infinite plate electrodes that sandwich an infinite slab of homogeneous material, this would result in an electric field magnitude of 1300 V cm−1, which is equivalent to the intended 1300 V cm−1 in the mice experiments (520 V for a 4 mm distance). This electric field magnitude has proven to be suitable for treating cutaneous tumors that can be properly placed between the two parallel plates (Belehradek et al 1993) (see also section 4). That is, it is large enough to guarantee that all cells in the region of interest undergo reversible electroporation but it is not large enough to cause significant irreversible electroporation. In the simulations, we define this magnitude as Etarget and we assume that the admissible tolerance range in the region to be treated is ±33%. In other words, we consider that successful electrochemotherapy is surely achieved in those regions in which E  0.66Etarget whereas the regions with E > 1.33Etarget are under risk of irreversible electroporation. In the resulting graphs, black color indicates E < 0.66Etarget; gray color 0.66Etarget  E  1.33Etarget; and white

Conductive gels for field homogenization in electroporation

6609

E > 1.33Etarget. Different gel conductivities ranging from 0 S m−1 (no gel) to 10 mS cm−1 were tried. Moreover, three cases in which the geometry of the gel is not optimal (see figures 5(a), (e) and (i)) were also analyzed. Because tissue conductivity depends on the electric field and the field in turn depends on the conductivity, some sort of iterative process is required to solve the resulting electric field distribution. Here, the model was solved as a sequence of steps in which the conductivity distribution for each step was defined by the electric distribution in the previous step (Pavselj et al 2005). Ten steps were employed here (but no significant changes were noticed after the sixth iteration). There is a significant number of studies on the small-signal passive electrical properties of different animal and human tissues (Grimnes and Martinsen 2000). However, when the electroporation phenomenon is involved then reliable data are extremely scarce (Pavselj et al 2005, Mossop et al 2006). Previously, we have produced some data on liver (Ivorra and Rubinsky 2007b), skeletal muscle (Ivorra et al 2007) and tumor (pending for submission). From those studies we know that tissue conductivity increases immediately after the onset of the electroporation pulse and that such conductivity raise increases with electric field magnitude until ‘saturation’ conductivity is reached. Pavselj et al (2005) did a similar FEM study to the one we report here and reported quite exhaustive experimental data on the conductivity of tissues under electroporation pulses. We have combined their experimental results with ours to create the conductivity models (conductivity versus electric field magnitude) for the simulations (figure 2). Quite arbitrarily we have chosen piecewise linear functions to approximate the conductivity behavior of tissues (σtumor (E > 100 V cm−1 ) = 1.5 mS cm−1 , σtumor (E =1600 V cm−1 ) = 2.6 mS cm−1 ; σmuscle (E > 200 V cm−1 ) = 1 mS cm−1 , σmuscle (E > 600 V cm−1 ) = 5 mS cm−1 ; σskin (E > 300 V cm−1 ) = 0.1 mS cm−1 , σskin (E > 1200 V cm−1 ) = 4 mS cm−1 ). Other functions, such as sigmoids could also be employed (Pavselj et al 2005); dispersion of measured values is too large in order to favor any specific sort of function. Nevertheless, we want to point out that we performed simulations with quite different models (some of them were far from fitting measured values) and results did not contradict the main conclusions of this study.

2.2. Conductivity measurement of some commercial gels It will be shown that a gel with a conductivity of about 5 mS cm−1 is a good choice for the technique described here. Such a gel can be easily implemented with agar and sodium chlorine (Ivorra and Rubinsky 2007a). Nevertheless, in view of the fact that the technique is quite tolerant to conductivity mismatches, we wondered whether there is a commercial gel with similar conductivity so that the technique would be adopted more easily in clinical environments. Gels intended for electrocardiography (ECG) or defibrillation are probably too conductive (conductivities larger than 10 mS cm−1) and, on the other hand, gels intended for ultrasonography do not need to conduct electricity and their conductivity could be extremely low. However, as manufacturers probably try to maximize the applications of their products, it is quite reasonable to assume that gels with intermediate properties exist. We have tried to locate data on electrical conductivity of gels used in electrocardiography (ECG) and ultrasonography. Unfortunately, the manufacturers of ECG gels provide little useful data and in the case of ultrasonography conductivity, data are almost non-existent. The scientific literature is also not very helpful in this case; we only found two quite old studies comparing the conductivity of different gels (Geddes 1972, Hummel et al 1989). Therefore, we collected some commercial gels and measured their conductivity.

A Ivorra et al

6610 (a)

(b)

(c)

Figure 2. Conductivity models (dark lines) employed in the simulations reported in this study: (a) tumor, (b) skeletal muscle and (c) skin. The crosses (×) represent actual measurements reported in Pavselj et al (2005) whereas the circles represent data obtained by us (partially reported in Ivorra et al (2007) for the muscle case).

The measurement cell consisted of two circular electrodes (1 cm diameter) on parallel plates separated by 2.2 mm. Cell constant (= 4.6 cm) was obtained by FEM simulations (Ivorra and Rubinsky 2007b) and verified with a saline solution NaCl 0.9% (resistivity at 25 ◦ C ≈ 70  cm). Measurements were performed with impedance analyzer 4294 A (Agilent, Inc.) at 10 kHz (oscillation voltage = 100 mV).

Conductive gels for field homogenization in electroporation

6611

For the animal experiments detailed in the next sections, we chose the Eko-Gel by CA.MI.NA, S.r.l. (Egna, Italy). The conductivity of this gel is 1.55 mS cm−1 and cannot be considered an optimal choice (see the next sections). The decision was motivated by the fact that this particular gel is already employed in clinical electrochemotherapy to improve the electrode–tissue electrical contact (despite the fact that this gel is intended for ultrasonography and not for electrical applications). 2.3. Animal model 2.3.1. Tumor cells culture and tumor production. Cells from a LPB cell line, a methylcholanthrene-induced C57 Bl/6 mouse sarcoma cell line (Belehradek et al 1972), were cultured using standard procedures in a minimum essential medium (Gibco BRL, CergyPontoise, France) supplemented with 100 U ml−1 penicillin, 100 mg ml−1 streptomycin (Sarbach, France) and 8% fetal calf serum (Gibco). C57 Bl/6 female mice, 6–8 weeks old, were inoculated subcutaneously in the left flank with 1 × 106 cells, producing in 9 to 10 days tumors of 3.9 ± 0.15 mm (mean ± standard deviation) diameter when they were subjected to electrochemotherapy. Animals were housed and handled according to the recommended guidelines (UKCCCR 1998). 2.3.2. Tumor treatment. At the start of the procedure, mice were anesthetized using a mixture of xylazine 12.5 mg kg−1 (Bayer Pharma, Puteaux, France) and ketamine 125 mg kg−1 (Parke Davis, Courbevoie, France). 10 μg of bleomycin (Roger Bellon S A, Neuilly-Sur-Seine, France) was injected intravenously (in the retro-orbital sinus) 4 min before the delivery of the electric pulses. Electroporation electrodes consisted of two stainless steel plates (10 mm width, 0.7 mm thick and 4 cm long). The tips of the electrodes were placed in direct contact with both sides of the cutaneous tumor, with the tumor between the parallel plates but without compressing it. The distance between the electrodes was 4 mm in all treatments. Conductive gel (Eko-Gel, CA.MI.NA, S.r.l., Egna, Italy) was applied according to the groups detailed in section 2.3.3. The square-wave electric pulses were generated by an electroporation power supply (CliniporatorTM, Igea, Carpi, Italy) able to apply high voltage pulses with a repetition frequency ranging from 1 Hz to 5 kHz. Here the treatment parameters were: eight pulses of 520 V with a duration 100 μs and repetition frequency of 5000 Hz. Maximum currents were recorded. Tumor volume was assessed every 3 days after treatment in order to evaluate response to the treatment. The formula used to compute the volume is V = ab2π /6, where a is the longest diameter of the tumor and b is the next longest diameter perpendicular to a. That is, tumors are approximated by ellipsoids. 2.3.3. Groups. Gel+ group (n = 9): the plate electrodes were placed on each side of the tumor, facing the largest diameter of the tumor, as represented in figure 1. The electrodes were embedded in the gel, and then another amount of gel was added in addition by using a syringe to completely cover the tumor and to remove possible air bubbles. Gel− group (n = 9): before applying the electrodes to the tumor, the electrodes were also embedded in the gel but they were shacked vigorously so that no excess of gel remained on the electrodes. After placement no additional gel was applied. The purpose of the thin layer of gel that remained on the electrodes was to improve the electrical contact between the electrodes and the tissues. Control group (n = 10): no treatment was applied to these tumors.

A Ivorra et al

6612 (a)

(b)

(c)

Figure 3. Electric field magnitude without gel when the applied voltage is 572 V (electrode separation distance = 4.4 mm). Pictures on the left show lateral cross sections (through tumor center and perpendicular to the skin and to the electrode plates) and pictures on the right show top cross sections at the tumor base. Simulation results for three different models are displayed: (a) homogenous and constant conductivity in all tissues (σ skin = σ muscle = σ tumor = 1 mS cm−1), (b) not uniform but constant conductivities (σ skin = 0.1 mS cm−1; σ muscle = 1 mS cm−1; σ tumor = 1.5 mS cm−1), (c) tissue conductivities according to models in figure 2. Black color indicates E < Emin (866 V cm−1); gray color Emin  E < Emax (ideal electric field magnitude for treatment, Etarget ± 33%); and white E  Emax (1733 V cm−1).

3. Results 3.1. Computer simulations Figure 3 shows simulation results without conductive gel (σ gel = 0 mS cm−1) for three different model constraints sets: (a) conductivity is uniform (σ skin = σ muscle = σ tumor) and it is not modified by the electric field (σ (E) = 1 mS cm−1), (b) conductivity is not uniform (σ skin = σ muscle = σ tumor) but it is constant (σ skin = 0.1 mS cm−1; σ muscle = 1 mS cm−1; σ tumor = 1.5 mS cm−1) and (c) conductivity for each tissue is according to the models depicted in figure 2 (not uniform and not constant). The three cases result in electric field distributions that are far from being homogeneous. Most tumor volume would not be treated properly (E < Emin, black color). We assume that the third set of constraints (used in following simulations) provides results closest to reality (Pavselj et al 2005). The second set (not uniform but constant conductivities) is clearly not realistic as it yields high electric fields in the skin but almost null in muscle and tumor; this even happens in simulations in which the tumor is properly compressed between the plates (simulations not reported here). Experimental results clearly

Conductive gels for field homogenization in electroporation

6613

(a)

(d)

(b)

(e)

(c)

(f)

Figure 4. Electric field magnitude when tumor is covered with gel. Different gel conductivities: (a) no gel (σ gel = 0 mS cm−1), (b) σ gel = 0.5 mS cm−1, (c) σ gel = 1 mS cm−1, (d) σ gel = 2 mS cm−1, (e) σ gel = 5 mS cm−1 and (f) σ gel = 10 mS cm−1. Black color indicates E < Emin (866 V cm−1); gray color Emin  E < Emax (ideal electric field magnitude for treatment, Etarget ± 33%); and white E  Emax (1733 V cm−1).

contradict such behavior: electroporation of inner tumors can be performed efficiently through the skin (Gothelf et al 2003). It is interesting to note that the first set of constraints (uniform and constant conductivities) provides results that are qualitatively similar to those from the third set, at least in this particular case. In fact, this can be explained quite intuitively: electroporation makes uniform tissue conductivity. In other words, tissues with higher resistivity (inverse of conductivity) experience higher electric fields which cause higher electroporation and larger resistivity drop. Figure 4 shows the simulated electric field magnitude distribution for different gel conductivities ranging from 0.5 mS cm−1 to 10 mS cm−1 (conductivities up to 50 mS cm−1 were tried but are not shown here). The ideal distribution would be Etarget ± 33% electric field magnitude (gray color) in tumor and null electric field in other tissue regions (electric field distribution in gel is not relevant). Such ideal distribution is best approximated for gel conductivities 2 mS cm−1 and 5 mS cm−1 (figures 4(d) and (e)). Lower conductivities result in non-treated areas (E < Emin V cm−1, i.e., black color) whereas higher conductivities cause unnecessary over-treated patches (white color). The former case must be avoided as any cancerous surviving cell can cause tumor recurrence; the consequences of the later case are not so grave in the case of tumor treatment. Because of this, we consider that the optimum gel conductivity range for the treatment of tumors by electrochemotherapy is between 5 mS cm−1 and 10 mS cm−1; lower gel conductivities can result in non-treated spots whereas higher gel

A Ivorra et al

6614 (a)

(e)

(b)

(c)

(f)

(i)

(g)

(j) (k)

(d)

(h)

(l)

Figure 5. Electric field magnitude when tumor is not properly covered with gel (σ gel = 5 mS cm−1). Model geometries are shown on the top (a, e, i). Result views: lateral cross section through tumor center and perpendicular to the skin and to the electrode plates (b, f, j); top cross section at tumor base (c, g, k) and front cross section through tumor center (d, h, l). Black color indicates E < Emin (866 V cm−1); gray color Emin  E < Emax (ideal electric field magnitude for treatment, Etarget ± 33%); and white E  Emax (1733 V cm−1).

conductivities will result in overtreatment of healthy areas and unnecessary high current drawn from the generator. Results shown in figure 4 are obtained with an optimum, and maybe unrealistic, geometry of the gel deposit. In figure 5 three different non-optimum gel geometries and their simulation results are shown. In these three cases, the conductivity of the gel is 5 mS cm−1. Results indicate that the technique is not only robust against conductivity mismatches but also against improper deposition of the gel (provided that a large enough amount of gel is employed). If the gel is applied as shown in figure 5(a) (uniform layer on the tumor surface), the technique will only provide enough field distribution improvement if the gel layer is very thick. In the particular case shown here (figure 5(a)), the gel layer was thick enough to cause reversible electroporation in the whole tumor. Nevertheless, this sort of gel deposition must be avoided; the top surface of the deposited gel must be as flat as possible. The second case (figure 5(e)) is probably the most realistic representation of what can actually happen when the gel is applied on the tumor; the gel spreads outside the tumor region and the gel surface on top of the tumor is slightly convex. We consider this case as non-optimum because the layer of gel on the tumor is very thin on top of the tumor (we recommend clinicians to use a much thicker layer of gel). Even so, results are similar to those obtained for the optimum gel geometry when the gel conductivity is 5 mS cm−1 (figure 4(d)). Finally, the third scenario (figure 5(i)) shows an extreme situation that should be avoided; the geometry along the axis between both electrodes must be as constant as possible. It was quite surprising for us that such irregular gel deposition was still able to produce reasonable results in terms of electric field distribution inside the tumor. 3.2. Commercial gel conductivities Table 1 contains the conductivities of the measured gels. None of them fits within the optimal range from 5 mS cm−1 to 10 mS cm−1. We will probably continue the search for a suitable commercial gel, but, even if such a gel is found, the implementation of a custom commercial gel for this specific application cannot be ruled out; from discussions with

Conductive gels for field homogenization in electroporation

6615

Figure 6. Tumor volume after treatment (mean ± standard deviation). Table 1. Measured conductivities of some commercial gels. Model

Supplier

Conductivity (mS cm−1)

UltraBio Sterile Eko-Gel Ultraphonic Aquasonic 100 Sterile Signal Gel Redux Gel

Sonotech, Inc. CA.MI.NA, S.r.l Pharmaceutical Innovations, Inc. Parker Laboratories, Inc. Parker Laboratories, Inc. Parker Laboratories, Inc.

0.0015 1.55 0.85 2.05 41.5 39.5

manufactures, we deduced that electrical conductivity is not a tightly controlled parameter in manufacturing, at least in the case of gels intended for ultrasonography. Therefore, there is the need for a conductive gel whose conductivity is specified, and guaranteed, from 5 mS cm−1 to 10 mS cm−1. Another desirable feature of such gel would be sterility. 3.3. Animal model Tumor growth results (figure 6) clearly indicate that the technique described here improves significantly the antitumor efficacy of electrochemotherapy (a Student’s t-test was applied on the samples of groups Gel+ and Gel− at day 18 after treatment and the obtained p-value was 7.3 × 10−5). The fact that complete regression is not achieved in the cases treated with gel (Gel+ group) may be due to the non-optimal conductivity of the employed gel (1.55 mS cm−1 instead of 5 mS cm−1) or to the selection of an insufficient voltage for electroporation. In any case, it is confirmed that it is better to use conductive gel in the way described here than just for electrode–tissue contact improvement (Gel− group). As expected, currents drawn from the generator in the Gel+ group (4.22 ± 0.66 A; mean ± standard deviation) were significantly larger (Student’s t-test p-value < 0.001) than currents drawn in the Gel− group (1.53 ± 1.10 A; mean ± standard deviation). 4. Discussion In a previous conference communication (Ivorra and Rubinsky 2007c), we showed, by means of a simplified finite element model, that mismatching errors between tissue conductivity and

6616

A Ivorra et al

gel conductivity were surprisingly well tolerated and that a proper conductivity for the gel would be 5 mS cm−1. Here we have employed an enhanced model that includes specific conductivities for skin, tumor and muscle tissues. The conclusions we reach are the same. Furthermore, here we demonstrate empirically that this technique increases the anti-tumor efficacy of bleomycin-based electrochemotherapy in the treatment of sarcomas inoculated into mice. Few days after inoculation, these tumors are soft and easily treated by compressing them between parallel plate electrodes. Complete tumor regression is easily achieved when the proper electrochemotherapy protocol is applied. However, here we wanted to mimic what happens in those tumors that cannot be compressed properly between two plates; a non-optimal electrode geometry was employed. In all the treatments, the electrode distance was 4 mm whereas the mean maximum tumor diameter was around 3.9 mm. Therefore, the tumor was not compressed properly and some dead spaces existed. Again we would like to clarify that the geometry we have employed here is not optimal and that, under normal conditions, clinicians will compress the tumor between the plates so that good contact and electric field uniformity are achieved (Corovic et al 2008). Tumor growth, days after treatment, shows significant improvement for those tumors in which conductive gel was employed to fill the space between both plate electrodes. Quite arbitrarily, we have chosen a tolerance of ±33% around Etarget for the definition of optimal field magnitude distribution. Here we briefly show that such tolerance interval is reasonable in the case of electrochemotherapy. From Al-Sakere et al (2007), we know that the protocol applied here (eight pulses of 100 μs) at 2000 V cm−1 induces some degree of irreversible electroporation in tumors (growth delay), but it is not sufficient to cause complete regression. On the other hand, in Pavselj et al (2005), in which electrochemotherapy efficiency is assessed indirectly by measuring cellular uptake of a radioisotope, it seems that tumor cells are not significantly destroyed by irreversible electroporation up to the highest field magnitude studied there (1350 V cm−1). Hence it seems that irreversible electroporation for tumors appears somewhere between 1350 V cm−1 and 2000 V cm−1. Since 1730 V cm−1 (Emax = Etarget + 33%) is close to the average of both points (1675 V cm−1), it can be considered a proper upper threshold for tumors. Moreover, we have performed a study (pending for submission) in which tumor impedance is measured after electroporation that seems to confirm that the irreversible electroporation threshold for tumors is of that order. The lower threshold (Emin = Etarget + 33% = 866 V cm−1) can also be justified from Pavselj et al (2005): there it is noticed that cellular uptake for sarcomas reaches a plateau, an indication of proper electroporation, around 800 V cm−1. A minor drawback of the technique presented here is an increase in the current drawn from the generator. In the particular experimental case presented here, which should be considered as a very extreme case, the average current for the Gel+ group almost tripled the average current for the Gel− group. On the other hand, it is interesting to note that the current drawn in the Gel+ group is more repetitive (smaller standard deviation) than the current drawn in the Gel− group. The large dispersion of current values in the Gel− group may result from the way in which the residual gel allows contact between the electrodes and the skin.

5. Conclusions We propose the use of conductive gels with an electrical conductivity equal to, or slightly larger than, that of tissues to fill dead spaces between plate electrodes gripping the tissue for electrochemotherapy of surface tumors. We have shown that this technique indeed improves the antitumor efficacy of electrochemotherapy in sarcomas implanted in mice. Furthermore,

Conductive gels for field homogenization in electroporation

6617

through finite element method simulations, we found that conductivity mismatching errors and gel deposition ‘errors’ are well tolerated by the technique. The technique is compatible with current in vivo electroporation methods, including electrochemotherapy, electrogenetherapy and irreversible electroporation ablation, based on plate electrodes or on needle electrodes. Our recommendations for researchers and clinicians who would like to try the use of large amounts of gel to increase the efficacy of the electrochemotherapy treatment for superficial tumors are the following: (1) use a transparent or translucent gel with an electrical conductivity between 5 mS cm−1 and 10 mS cm−1, (2) embed the electrode plates in the gel before placement, (3) after electrode placement, apply a generous amount of gel so that the tumor is completely covered, (4) remove any possible air bubbles and (5) apply the pulses. It is advisable to monitor the applied voltage (e.g. by means of an oscilloscope) in order to be sure that the generator was able to deliver the extra amount of current because of the gel. Acknowledgment This work was supported by grants of CNRS and IGR and by the US National Institutes of Health (NIH) under grant NIH R01 RR018961. The staff of the Service Commun d’Exp´erimentation Animale (headed by Dr P Gonin) of the Institut Gustave-Roussy is also acknowledged for mice housing. We want to express our gratitude to Parker Laboratories, Pharmaceutical Innovations and Sonotech for providing us with free gel samples. BR has a financial interest in Excellin Life Sciences and Oncobionic, which are companies in the field of electrical impedance tomography of electroporation and irreversible electroporation, respectively. References Al-Sakere B, Andr´e F, Bernat C, Connault E, Opolon P, Davalos R V, Rubinsky B and Mir L M 2007 Tumor ablation with irreversible electroporation PLoS ONE 2 e1135 Belehradek J Jr, Barski G and Thonier M 1972 Evolution of cell-mediated antitumor immunity in mice bearing a syngeneic chemically induced tumor. Influence of tumor growth, surgical removal and treatment with irradiated tumor cells Int. J. Cancer 9 461–9 Belehradek M, Domenge C, Luboinski B, Orlowski S, Belehradek J J and Mir L M 1993 Electrochemotherapy, a new antitumor treatment. First clinical phase I-II trial Cancer 72 3694–300 Corovic S, Al-Sakere B, Haddad V, Miklavcic D and Mir L M 2008 Importance of contact surface between electrodes and treated tissue in electrochemotherapy Technol. Cancer Res. Treat. 7 393–400 Davalos R V, Mir L M and Rubinsky B 2005 Tissue ablation with irreversible electroporation Ann. Biomed. Eng. 33 223 Dean D A 2005 Nonviral gene transfer to skeletal, smooth, and cardiac muscle in living animals Am. J. Physiol. Cell Physiol. 289 C233–45 Dev S B, Dhar D and Krassowska W 2003 Electric field of a six-needle array electrode used in drug and DNA delivery in vivo: analytical versus numerical solution IEEE Trans. Biomed. Eng. 50 1296 Edd J, Horowitz L, Davalos R V, Mir L M and Rubinsky B 2006 In-vivo results of a new focal tissue ablation technique: irreversible electroporation IEEE Trans. Biomed. Eng. 53 1409–15 Edd J F and Davalos R V 2007 Mathematical modeling of irreversible electroporation for treatment planning Technol. Cancer Res. Treat. 6 275–86 Geddes L A 1972 Surface electrodes Electrodes and the Measurement of Bioelectric Events (New York: WileyInterscience) pp 45–106 chapter 2 Gehl J, Sorensen T H, Nielsen K, Raskmark P, Nielsen S L, Skovsgaard T and Mir L M 1999 In vivo electroporation of skeletal muscle: threshold, efficacy and relation to electric field distribution Biochim. Biophys. Acta 1428 233–40 Gothelf A, Mir L M and Gehl J 2003 Electrochemotherapy: results of cancer treatment using enhanced delivery of bleomycin by electroporation Cancer Treat. Rev. 29 371–87

6618

A Ivorra et al

Grimnes S and Martinsen O G 2000 Bioimpedance and Bioelectricity Basics (London, UK: Academic) Hofmann G A 2000 Instrumentation and electrodes for in vivo electroporation IN Electrochemotherapy, Electrogenetherapy and Transdermal Drug Delivery: Electrically Mediated Delivery of Molecules to Cells ed M J Jaroszeski, R Heller and R A Gilbert (Totowa, NJ: Humana Press) pp 37–61 Hummel R S III, Ornato J P and Clarke A M 1989 The misuse of high impedance electrode gels for defibrillation: apotential fire hazard Proc. Annu. Int. Conf. IEEE Eng. Med. Biol. Soc. (Seattle, WA: IEEE) pp 1547–8 Ivorra A, Miller L and Rubinsky B 2007 Electrical impedance measurements during electroporation of rat liver and muscle 13th International Conference on Electrical Bioimpedance ed H Scharfetter and R Merva (Berlin: Springer) pp 130–3 Ivorra A and Rubinsky B 2007a Electric field modulation in tissue electroporation with electrolytic and non-electrolytic additives Bioelectrochemistry 70 551–60 Ivorra A and Rubinsky B 2007b In vivo electrical impedance measurements during and after electroporation of rat liver Bioelectrochemistry 70 287–95 Ivorra A and Rubinsky B 2007c Optimum conductivity of gels for electric field homogenization in tissue electroporation therapies IV Latin American Congress on Biomedical Engineering, Bioengineering Solutions for Latin America Health ed C M¨uller-Karger, S Wong and A La Cruz (Berlin: Springer) pp 619–22 Jaroszeski M J, Heller R and Gilbert R 2000 Electrochemotherapy, Electrogenetherapy, and Transdermal Drug Delivery: Electrically Mediated Delivery of Molecules to Cells (Totowa, NJ: Humana Press) Marty M et al 2006 Electrochemotherapy—an easy, highly effective and safe treatment of cutaneous and subcutaneous metastases: results of ESOPE (European Standard Operating Procedures of Electrochemotherapy) study Eur. J. Cancer Suppl. 4 3–13 Miklavcic D, Beravs K, Semrov D, Cemazar M, Demsar F and Sersa G 1998 The importance of electric field distribution for effective in vivo electroporation of tissues Biophys. J. 74 2152–8 Miklavcic D, Corovic S, Pucihar G and Pavselj N 2006 Importance of tumour coverage by sufficiently high local electric field for effective electrochemotherapy Eur. J. Cancer Suppl. 4 45–51 Miklavcic D, Semrov D, Mekid H and Mir L M 2000 A validated model of in vivo electric field distribution in tissues for electrochemotherapy and for DNA electrotransfer for gene therapy Biochim. Biophys. Acta 1523 73–83 Mir L M 2000 Therapeutic perspectives of in vivo cell electropermeabilization Bioelectrochemistry 53 1–10 Mir L M, Gehl J, Sersa G, Collins C G, Garbay J-R, Billard V, Geertsen P F, Rudolf Z, O’sullivan G C and Marty M 2006 Standard operating procedures of the electrochemotherapy: instructions for the use of bleomycin or cisplatin administered either systemically or locally and electric pulses delivered by the CliniporatorTM by means of invasive or non-invasive electrodes Eur. J. Cancer Suppl. 4 14–25 Mir L M, Moller P H, Andre F and Gehl J 2005 Electric pulse-mediated gene delivery to various animal tissues Advances in Genetics (New York: Academic) pp 83–114 Mossop B J, Barr R C, Henshaw J W, Zaharoff D A and Yuan F 2006 Electric fields in tumors exposed to external voltage sources: implication for electric field-mediated drug and gene delivery Ann. Biomed. Eng. 34 1564–72 Pavselj N, Bregar Z, Cukjati D, Batiuskaite D, Mir L M and Miklavcic D 2005 The course of tissue permeabilization studied on a mathematical model of a subcutaneous tumor in small animals IEEE Trans. Biomed. Eng. 52 1373 Rubinsky B 2007 Irreversible electroporation in medicine Technol. Cancer Res. Treat. 6 255–60 Sel D, Cukjati D, Batiuskaite D, Slivnik T, Mir L M and Miklavcic D 2005 Sequential finite element model of tissue electropermeabilization IEEE Trans. Biomed. Eng. 52 816–27 Sel D, Mazeres S, Teissie J and Miklavcic D 2003 Finite-element modeling of needle electrodes in tissue from the perspective of frequent model computation IEEE Trans. Biomed. Eng. 50 1221 Sersa G 2006 The state-of-the-art of electrochemotherapy before the ESOPE study; advantages and clinical uses Eur. J. Cancer Suppl. 4 52–9 Sugibayashi K, Yoshida M, Mori K, Watanabe T and Hasegawa T 2001 Electric field analysis on the improved skin concentration of benzoate by electroporation Int. J. Pharm. 219 107–12 UKCCCR 1998 United Kingdom Co-ordinating Committee on Cancer Research (UKCCCR) Guidelines for the Welfare of Animals in Experimental Neoplasia (Second Edition) Br. J. Cancer 77 1–10 Valic B, Pavlin M and Miklavcic D 2004 The effect of resting transmembrane voltage on cell electropermeabilization: a numerical analysis Bioelectrochemistry 63 311–5 Zupanic A, Corovic S and Miklavcic D 2008 Optimization of electrode position and electric pulse amplitude in electrochemotherapy Radiol. Oncol. 42 93–101