Generation and measurement of low-temperature plasma for cancer therapy: a historical review

Abstract

This review provides a description of the historical background of the development of biological applications of low-temperature plasmas. The generation of plasma, methods and devices, plasma sources, and measurements of plasma properties, such as electron dynamics and chemical species generation in both gaseous and aqueous phases, were assessed. Currently, direct irradiation methods for plasma discharges contacting biological surfaces, such as the skin and teeth, are related to plasma biological interactions. Indirect methods using plasma-treated liquids are based on plasma–liquid interactions. The use of these two methods is rapidly increasing in preclinical studies and cancer therapy. The authors address the prospects for further developments in cancer therapeutic applications by understanding the interactions between the plasma and living organisms.

Keywords:

• Low-temperature plasma (LTP)
• reactive oxygen and nitrogen species (RONS)

Introduction

The universe is 99% plasma, except for the unresolved dark matter. As an example, the temperature of the Sun reaches as high as 15.7 million °C in its inner core, ionizing all the matter containing hydrogen and helium. The ionized state of gas is called the plasma state. The astrophysical phenomena near the Earth can be regarded as matter behaving in the plasma state, as the universe consists almost entirely of plasma [1]. A stream of charged particles is ejected from the solar surface corona to the planets, including the Earth. This stream, called the solar wind, is deflected by the Earth’s magnetic field and protects the Earth from solar wind irradiation, flowing into Earth’s poles, ionizing the atmosphere, and appearing as visible auroras in the sky [2,3]. Plasma is considered as the fourth state of matter, in which positively and negatively charged particles are quasi-neutralized, occurring naturally at temperatures higher than that of the gaseous phase. Plasmas occurring in nature span several orders of magnitude in terms of density and temperature (Figure 1). Like lightning, the disruption in the insulation by the flow of a large electrical current produces a plasma state called an arc discharge, which was discovered in the 1800s and cannot be touched because of its high temperature.

Figure 1. Logarithmic diagram of plasmas in density and temperature; T_e, T_i, and T_g are the temperatures of electrons, ions, and gases, respectively (thermal plasmas are in an equilibrium state of T_e, T_i, and T_g; contrarily, the low temperature or non-thermal or cold plasmas are in a non-equilibrium state among T_e, T_i, and T_g. that electrons are differently heated up; the non-equilibrium atmospheric pressure plasma (NEAPP) source can prepare plasmas at ambient temperature under atmospheric pressure for plasma-biological treatments of living organisms and liquids).

A sophisticated method has been developed recently to use such plasma states in electrical discharges, with various applications, such as the lighting of fluorescent lamps, material synthesis of sputtering deposition, and dry etching [4]. With the expansion of its applications and bio-applications, including plasma medicine, it has attracted significant attention.

Plasma medical science is a novel academic field, which aims to elucidate the mechanisms of physiological outputs induced by gas plasma [5]. Importantly, low-temperature plasmas are extremely relevant compared with the arc discharge plasmas, in which the gas is equalized to an electron temperature of ∼10,000 K [6]. The development of this low-temperature plasma was based on a study on electrostatics and micro-plasmas [7]. Currently, plasma is generated at body temperature under atmospheric pressure [8], because it is essential for the development of plasma-based cancer therapy [9–11].

After the success of the artificial generation of a glow-like plasma around an ambient temperature under atmospheric pressure, and the number density of electrons in the plasma approaching 10¹⁶ cm⁻³ [12], it is only 10 years. In 2018, a book titled “Plasma Medical Science” interpreted complex biological responses resulting from the interaction of biological material with low-temperature plasma, similar to cold plasma or non-thermal plasma [13]. Since then, the latest developments in plasma-based cancer therapy have been summarized.

This review describes the history of plasma medical science and technologies for the generation and measurement of low-temperature plasmas. First, the historical background of low-temperature plasma cancer therapy is briefly described. Second, the technological views of the generation methods are described. Herein, plasma sources are categorized into three types: corona and streamer discharges, dielectric barrier discharges, and remote or effluent plasmas. Third, to evaluate the therapeutic efficiency and efficacy, quantitative measurements of plasma were performed, which involved electron dynamics and chemical species in the gaseous and aqueous phases. Finally, the progress in biological assessments for clinical tests of plasma cancer therapy is reviewed.

Generations of low-temperature plasma

Herein, recent developments in sources of low-temperature plasma generation are reviewed. Based on spatiotemporal analyses of inception with high spatial and temporal resolution, significant results have been achieved.

The low-temperature plasma sources are categorized into three types [14]: (i) coronal or spark discharges between two conductive electrodes, (ii) dielectric barrier discharges (DBD) with insertion of dielectrics between two electrodes, and (iii) effluent plasma discharge region or remote treatments (Figure 2).

Figure 2. Schematics of plasma sources and configurations of treatments (DBD stands for dielectric barrier discharge) [14] (Reprinted from Jpn J Appl Phys 61, SA0805 (2022)).

Streamers of molecular transduction into cells

In Type (i), Jinno et al. recently developed a micro-discharge system for plasma gene transfection and introduced certain molecules into biological cells [15]. Electroporation using high-voltage electrical pulses was first described by Neumann in 1982 [16]. Previously, different types of plasma sources have been compared. To date, the micro discharge plasma, in which a 70-µm diameter electrode is used, achieved high efficiencies simultaneously both transfection efficiency and cell viability while a short period of treatments is less than 0.01 s [17] (Figure 3). The reasons for these simultaneous efficiencies are considered a complex mixture of electrical, chemical, and biochemical effects, including depolarization owing to the electrical charges, pore formation, peroxidation of cell membranes, and endocytosis enhanced by cell activation [19–21]. They also analyzed the electrical current flow in cells in a Petri dish during an electrical pulse supply [18,22]. Accordingly, the small size of the plasma generated is effective in reducing cell damage. This enables the introduction of genes or molecules into living plants and fish eggs [15]. The regulation of plasma volume and treatment time rescued by oxidative damage, such as the formation of 8-oxoguanine (8-oxoG) on nuclear and mitochondrial DNA in A549 human lung cancer cells [23]. When plasma irradiation was carried out on Arabidopsis thaliana seedlings, growth retardation was observed [24]. Plasma treatments affect the differentiation and morphology of human induced pluripotent stem cells (hiPSCs); however, they do not appear to cause extensive DNA damage [25]. In summary, Jinno stated that the micro-discharge plasma is suitable for applying both electrical and chemical stimuli. The micro-discharge plasma method enables the use of small or fewer stimuli without causing cellular damage.

Figure 3. Plasma streamer in a gene transfection device using microdischarge plasma [18] (Reprinted from PLOS One 16, e0245654 (2021)).

Streamer generation

The aforementioned streamers and coronal sparks are typically operated using short electrical pulses. Generally, cosmic rays, radiative emissions from solids, and even ultraviolet irradiation produce electrons in space. In addition to the presence of an electric field in the space-accelerating electrons, the electric field is reconstructed by the creation of electric charges. The phenomenon of gaseous electric discharge has been studied extensively. Based on the contributions of previous studies, the discharges were modeled using the Townsend theory and Meek and Loeb streamer theory [6]. However, there are many unknown phenomena regarding the actual electric field of plasma, and many systems remain too complex to understand if the boundary between the plasma and liquids is considered in addition to the other phases and chemical reactions.

Owing to the increased sophistication of measurements, the inception phenomenon of plasma discharges has been used to determine the spatiotemporal development of the electric field and electron density. When a high voltage is applied to an electrode, electrons multiply through ionization owing to highly energetic electron collisions and/or photoionization. Strong electric fields accelerate the appearance of background electrons as energetic electrons. With respect to the electron density, the discharge regimes change with the changing electrical current behavior. The inception regimes have developed in the order of corona, glow, spark, corona, and horizontal ellipsis. The initial corona grows glowing and then sparks. The spark regime is interrupted by the growth of the discharge owing to the short pulse duration. This interruption results in instability of the discharge; however, oxygen addition boosts the duration of the spark regime [26]. Accordingly, the regimes can be transited within nanoseconds, and the memory effect, in which the electrons stay on the spark or streamer channel, enhances the transitions.

The experiments were conducted with airflow on the positive streamers. As the airflow speed increased, the streamers propagated upward. All the parameters of the inter-pulse period, peak current, deposited energy, and pulse width with respect to current pulsation were dispersed in more bursts of high-energy streamers with air flow speed [27].

In computational studies, the streamer propagation of a helium plasma jet ejected into the ambient air was calculated. The streamer velocity and spatiotemporal profiles of the plasma parameters were obtained for various helium flow rates [28]. Meanwhile, ionization wave inception and propagation channel breakdown were calculated, considering the information on the streamer structure, propagation velocity, radius, and parameters of the streamer plasma (electron density in the streamer channel and peak electric field in the streamer head) in various media [29]. The generation of wide negative streamers in air and helium has been demonstrated in a simulation [30]. The ionization and Penning reactions in streamer formation in long gaps have been numerically studied [31]. Non-axisymmetric simulations of negative and positive streamers were developed, and the main characteristics of the positive and negative streamers, including the morphology, distribution pattern of space charges, local electric field, diameter, length, and velocity, were reported [32,33].

The positive and negative streamers were calculated using 2D axisymmetric fluid simulations. Steady negative streamers can continue propagating over tens of centimeters; however, the properties of positive streamers are significantly different [34]. Faster positive streamers can propagate in significantly lower background fields owing to the reduced electron attachment and recombination [35]. The positive streamer discharges in air at 100 mbar were compared via experiments and calculations using a 2D axisymmetric drift-diffusion-reaction fluid model [36]. The agreement between the fluid and particle simulations was examined [37]. Transport data, background ionization level, photoionization rate, gas temperature, voltage rise time, and voltage boundary conditions were considered; however, an increase in gas temperature during operation was a major source of discrepancy [36].

The waveforms of the nanosecond voltage source were controlled using a solid-state impedance-matched Marx generator. The stepped waveform can be used to control streamer propagation [38].

Streamers simulations

The streamers under atmospheric pressure in the Ar gas were calculated using a 2D fluid model. Electrical breakdown forms primary and secondary streamers, and parameters, such as the polarity of the applied voltage affect propagation properties, including the propagation speed and streamer head size. For positive streamers, a low-voltage operation is useful for stable high-electron-density discharges under atmospheric-pressure argon [39] (Figure 4).

Figure 4. Example of simulation of streamer: (top) calculation setup and conditions, (bottom) spatiotemporal behavior of electron production rate in atmospheric pressure argon discharge when a positive DC voltage of +10 kV is applied to the pin electrode (a region of high electron production arising from the pin electrode shows the generation of a secondary streamer; a similar high-value region from the plane electrode shows a return stroke) [39] (Reprinted from J Phys D 53, 265204 (2020)).

Streamers propagate ionization fronts with self-organized field enhancements at their tips. The chemistry of streamers in their two main stages, inception, and propagation, has been described [40]. The interval between the pulse discharges was examined using calculations. The corona discharge moved toward the streamer discharge in the air. No difference between the streamers with and without a preceding corona discharge was observed because of the primarily positive ions [41]. During the early stages of the electric breakdown of gases during lightning, electron attachment by the electronegative gas SF₆ was tested to determine the effects of the propagation of positive streamers [42]. Positive streamers in ambient air were calculated using a fluid simulation [43]. The streamer discharges with airflow was calculated using 3D fluid models [44]. Positive streamers at an air-gap distance of up to 3 m have been calculated [45]. Before the inception of a stable leader, photoionization was performed via external re-illumination and observed through Schlieren imaging [46]. The ionization wave in pure nitrogen at 27 mbar was calculated using 2D fluid simulations [47]. Positive and negative streamers in air were observed and calculated using the 2D fluid model. Runaway electrons, which have been discussed in particle-in-cell Monte Carlo collision simulations [48], stop the propagation of positive streamers [49].

There are two categories of simulation methods for electric fields: finite and boundary elements. Although most calculations use the former approach, there have been reports on the use of the latter. Streamers were calculated using a spectral element method based on the Galerkin method with the hierarchical Poincare Stekloy scheme [50]. The electric field was calculated using the charge simulation technique [51]. Streamer simulations based on an electron source ahead of the ionization front have been well-developed. Multiple streamer branches should be calculated using three-dimensional structural simulations [52]. Despite the fact that many phenomena cannot be reproduced, universal phenomena are being analyzed both computationally and experimentally.

Surface discharges on dielectrics or in liquids

The surface DBD was calculated using 2D particles in cell simulations [53]. The segmentation of the electrode was argued, and the arrangement of the electrode geometry improved the plasma performance [54].

Streamers in liquids have traditionally been considered gas–phase ignition mechanisms, in which gas bubbles are formed before ignition [55]. According to these observations, the behavior of propagation channels in a liquid can be reproduced using a simulation model of the transport of hydrated electrons. This model suggests that local ionization rates at the ionization front dominate the propagation with channel formation in the liquid rather than advection or diffusion of electrons in the gas phase [56]. Furthermore, in liquid nitrogen, streamer-induced cryoplasmas showed bush-like streamers rather than filamentary streamers [57]. Originally, helium discharges at liquid-nitrogen temperature of 77 K were demonstrated [58]. Furthermore, interesting plasma phenomena under cryogenic conditions that occur in high-density media at low temperatures have been reported [59–62].

Dielectric barrier discharge (DBD) jet

In Type (ii), researchers have used various geometrical designs for sources because of their ease of use. Recent reviews have focused on their applications in cancer therapy [63–66]. Modeling of plasma–liquid interactions has been demonstrated [67,68] and the effects of electromagnetic fields from plasma bullets have been discussed [65]. Martinez et al. simulated the interactions between plasma and dielectrics. Cancer cells were modeled using dielectric materials, and their permittivity could modify the polarization and surface potentials. Based on this, they proposed that local electric fields may help physically target specific cells [69].

Generally, the gas temperature of atmospheric-pressure plasma jets tends to increase with discharge time, even when helium is flowing. As an example of the growth condition of fission yeast, the temperature must be ∼30 °C. To avoid thermal damage, the Peltier device was used to control the feeding gas temperature, which was maintained at an almost constant value [70]. In other reports, the plasma gas temperature was controlled by delivering a heater control of the precooled gas to the plasma source [71,72]. Essentially, the importance of gas temperature control should be addressed when living organisms are irradiated with discharge plasma [70].

Surface DBD includes a type of twin electrode operated using nanosecond and microsecond voltage sources. The two microsecond operations formed filamentary mode and the one nanosecond operation generated homogeneous mode without filamentary discharges [73].

Coaxial dielectric barrier helium discharges were calculated using an axisymmetric fluid model [74]. Their computational results tailored the three ignition stages of the plasma generation inside the discharge tube. First, a Townsend-glow-type plasma spreads in the region between the electrodes. Second, a plasma streamer propagates axially. Third, the bullet transitions into a surface discharge at the dielectric surface. The streamer propagates through the trapping of charges with strong electric fields induced by the grounded electrode underneath the dielectric barrier and by the surface charge accumulated on the dielectric surface [74] (Figure 5).

Figure 5. Three stages of plasma electron generation: (a) first stage (Townsend-glow type discharge), (b) second stage (streamer or bullet type discharge), and (c) third stage (surface type discharge) (reduced electric field (upper row) and electron number density (lower row) in the gas are shown; the first stage of the nearly uniform distributions of an electric field wavefront (streamer head) moves to the second stage of growth of electron number and to the third stage with high density near the dielectric surface) [74] (Reprinted from Appl Phys Express 13, 086001 (2020)).

Furthermore, the magnetic fields are significant. When the helium-based atmospheric-pressure plasma jet was operated with an axial magnetic-field of 0.587 T, electron density was ∼2.4 times higher at 1.2 × 10¹⁴ cm⁻³ compared to that with no magnetic field application [75]. Moreover, by applying an external magnetic field longitudinally, self-focusing of the discharge was studied using numerical simulations. The ionization wave decelerates radially, which leads to a change in the electron energy distribution function, decrease in the average electron energy, rate of gas ionization, and electron mobility in crossed electric and magnetic fields [76].

Multiple jet devices were introduced by the Orlean group and constructed in a collection of individually operated 52 jets in a circular area of diameter 50 mm, and are used in the therapeutic decontamination of chronic wounds. A pilot study in a hospital showed the inactivation of Staphylococcus aureus, Pseudomonas aeruginosa, and Escherichia coli [77]. Because the scaling of plasma generation is difficult, the aggregation of individual sources is an easier solution.

When both large and small areas are to be treated, the problem is not solved for the treatment of large areas; it is still difficult to achieve a homogeneous discharge if spark or DBD types are used.

Effluent of discharges

In Type (iii), a high voltage with high frequency in the MHz range was applied to generate plasma discharge inside the enclosure of flowing argon gas. One example is the kINPen MED (Neoplas Tool GmbH, Germany) [78] and another example is the gliding arc device. A high-current arc thermalizes gases similar to the electron temperature and is typically greater than 10,000 K. In contrast, a low-current gliding arc was generated between the two orifice electrodes and the gas temperature decreased to below 1000 K. To improve the capacity, multiple electrodes in a reactor have been examined [79]. In 2010, a magnetically stabilized gliding arc device consisting of a cylindrical reactor with concentric electrodes made of stainless steel was developed [80,81]. In contrast, a reverse vortex gliding design was developed at the Drexel University, and at present, a 100 L capacity plasma system with a submerged 3 kW gliding arc has been scaled up [82].

A stable plasma jet of diameter 20 mm was generated by swirling the discharges toward the tube nozzle. Metastable helium plays an essential role in ionization via the Penning effect. Therefore, trace impurities, such as N₂, O₂, or H₂O, decrease the pre-ionization level and impede the homogeneous discharges [83]. The scale-up of the helium plasma jet was achieved by incorporating a porous ceramic plate into the nozzle [84].

Non-equilibrium plasma was originally reported by enclosing the discharge with dielectric walls, where the dielectric walls sandwiched the metal electrodes [12]. This source was first used for cancer treatment in 2012 [85]. The experiments were carried out by irradiating the cells attached to the bottom of the Petri dishes containing the cell culture medium. The plasma was in contact with only the liquids, and biological effects were observed with the use of the plasma-treated liquids. Thus, the plasma can be applied indirectly in this case. Plasma-irradiated liquid solutions, such as the plasma-activated culture medium (PAM), are used to selectively kill the cancer cells without killing the normal cells [86,87].

Recently, Nakamura reported results obtained from a new plasma-treated liquid production system [88]. If the plasma is irradiated under atmospheric air, it entrains the surrounding air, and rich reactive oxygen and nitrogen species (RONS) are generated; however, the reproducibility of the concentrations of chemical species fluctuates because of the lack of control of the surrounding environment. To satisfy these requirements, the plasma source is mounted in an enclosure that can be purged with pure gas, such as pure argon or pure nitrogen (Figure 6). According to a report by Nakamura, they purged the enclosure with pure argon gas [10 standard liters per minute (slm)] and flowed a mixture of 10% pure nitrogen, 10% pure oxygen, and balanced pure argon gases [88]. Using this system, the mixing ratios of nitrogen and/or oxygen were controlled to achieve selective killing performance. Consequently, a plasma source with excellent reproducibility was developed.

Figure 6. Non-equilibrium atmospheric pressure plasma system (before the preparation of the plasma-activated solution, the chamber is filled with Ar gas, which flows into three lines of the plasma head, whereas the reactive gases flow only into the center of the three lines) [88] (Reprinted from Plasma Process Polym 17, 1900259 (2020)).

In addition, the plasma effluent source contains all these effects, including the charged species, electronically neutral atoms and molecules (neutrals), radicals, and light. A radical source is installed to eliminate the charged species and lights [89–97]. The remote part of the plasma source is supplied with only electronically neutral species and radicals.

Historic overview of generations of low-temperature plasma for bioapplications

Overall, low-temperature plasma devices have significantly progressed over the past decade (Figure 7). Matching was developed to track both experiments and simulations. As described later, plasma generation, rather than direct plasma treatments by touching plasma on biological surfaces, and indirect plasma treatments with the preparation of plasma-treated solutions, have been used in various applications. Surprisingly, the visible plasma could be treated with irradiated solutions without directly touching the samples. As a result, the discipline of the interactions between the plasma and liquid deepened. Electrosurgical devices and endoscopic instruments, such as plasma scalpels and argon plasma coagulators have been introduced as surgical instruments. They typically operate in high-temperature plasma at high gas temperatures. The development of low-temperature plasma with a low gas temperature has progressed in recent years because of its ability to produce effects different from those at such high temperatures. Methods have been developed for use on solid surfaces in dermatology and dentistry. In other methods, treatments of drugs and solutions for use in surgery, as well as liquid applications, have been developed to understand plasma and liquid interactions. The former is called “direct irradiation” and the latter is called “indirect irradiation.” Under these circumstances, the control of the plasma treatments or processes is the next issue to be addressed.

Figure 7. Timeline of the development of representative plasma devices for medical treatments DBD: dielectric barrier discharge, FE-DBD: floating electrode DBD, NEAPP: non-equilibrium atmospheric pressure plasma, LF-jet: low frequency operated jet. Taylor [98], link, Glover [99], Storek [100], Stoffels [101], Kalghatgi [102], Shimizu [103], von Woedtke [104], Kitano [105], Ikehara [106], Kanazawa [7], Neumann [16], Iwasaki [12], and Nakamura [88].

Measurements of low-temperature plasma

To evaluate the plasma processes, a hierarchical structure must be considered. In other words, to obtain results after plasma treatment, a cascade of plasma generation, chemical reactions in the gas phase, transport toward the boundary, aqueous chemistry, and/or biochemical action in living organisms must be considered [107] (Figure 8).

Figure 8. Schematic of hierarchical analysis of spatiotemporal measurements. LTS: laser Thomson scattering; E-FISH: electric field induced second harmonic generation; LIF: laser induced fluorescence; ESR: electron spin resonance; LC-MS: liquid chromatography mass spectrometry [107] (Modified from Jpn J Appl Phys 61, SA0805 (2022)).

Measurements of physical properties of plasma

Optical emission spectroscopy (OES) is widely used for plasma diagnosis. As previously described, electron density is a key parameter of plasma properties. For this purpose, the Stark broadening of hydrogen and helium emissions can be used. Furthermore, Stark polarization spectroscopy of helium emission (He I line) at a wavelength of 492.19 nm in nanosecond streamer plasma at helium pressure of ∼16 kPa (120 Torr) has been reported [108]. The surface discharge distributions of the helium nanosecond plasma jet in ambient air were observed through two-dimensional images [109]. The spatiotemporal electric fields of the helium DBD jet were measured using Stark polarization spectroscopy combined with the high-spatial-resolution monochromatic imaging technique, and the electric field values in the range of 4–25 kV/cm were obtained [110].

Positive streamers in the air propagating along non-planar dielectric plate surfaces were observed and compared with the results of two-dimensional simulations [111]. Surface streamers with high local electric fields have been observed [112]. In other words, surface charging significantly affects the development of local electric fields.

On dielectric surfaces, non-uniform irradiation of plasma is often observed. Self-organized patterns have been observed on the anode liquid surface during the irradiation of atmospheric glow discharge [113]. A star-shaped pattern of microchannels was formed on the gelatin matrix after plasma irradiation [114]. Petal-like patterns were observed in the argon-plasma jets. The DBD operation involved two positive and one negative discharge cycles. The second positive discharge exhibited complex dispersed patterns of stripe homogeneity and branching [115]. Surface discharges in distilled water were analyzed. In DBD operations, the stochastic nature of the plasma charges on the water surface has been experimentally observed [116]. After jet irradiation, surface charges accumulate. The actual charge accumulation depends on the humidity, frequency, and voltage [117].

Positive streamers in the long-distance gap in the range of 9–12 cm were calculated using 2D simulations with changing pressures and humidity. Increasing the water vapor content hampers the streamer development owing to the dissociative recombination of electrons with electron attachment to O₂ molecules [118].

The optical emissions from diffuse discharges of argon and hydrogen were measured. The intense emission arising from the Ar dimers (Ar₂*) at 126 nm decreased in intensity owing to the addition of small xenon to the other band arising from dimers ArXe* and Xe₂*. In pure hydrogen, H₂* at 160 nm had the highest intensity, and the addition of Ar to the H₂ changed Ar⁺ ion at 191 nm, and the broad band in the range 220–280 nm appeared again. These behaviors were determined by the changes in the excitation temperature and electron density, and could be diagnosed from the emission spectra [119].

The theory of inception based on lightning was examined by observing the high-speed emissions. The small particles are enhanced to form streamers [120]. In multiple jets, the analysis showed that multiple corona discharge inceptions were unable to shield the electrostatic interactions among the emitters [121]. Discharge inception was analyzed via Schlieren imaging [122] and high-speed optical emission spectroscopy [123].

The Adamovich group reported that the spatial resolution of electric field measurements was significantly improved using the electric field-induced second harmonic (EFISH) method [124–126]. The EFISH experimental data were examined using two-dimensional axisymmetric numerical fluid simulations of nanosecond discharges in the air [127,128].

The axial electric field in the nanosecond DBD jet of argon flow was measured using EFISH. Before arriving at the primary ionization wavefront, the radial distribution of the electric field peaks at the center, the electric field expands radially and then decreases. After passing through the primary ionization wave, the electric field is hollowed into a radial distribution [129].

The EFISH experimental data were examined by numerically simulating the fluid model. The drift–diffusion approximation of diffusive ionization waves and fast ionization waves can be applied to calculate the local electric fields. Electric fields evolve during the inception, propagation, and channel breakdown stages. Helmholtz photoionization is represented by the parameters N₂, O₂, CO₂, and air [130].

When the DBD jet was operated in different conditions, such as positive and negative polarities, oxygen or nitrogen impurity, voltage raising time was in the range of 60–200 µs and peak electric fields were affected by the conditions [131].

EFISH measurements of the electric field generation in spark discharges between the two electrodes were performed [132], and the calibration methods for the obtained data were reviewed [133]. In addition, a piezoelectric transformer can be used to calibrate the electric fields in open air [134]. Using a spherical lens, one dimensional electric fields of the primary to secondary transition of a single filament positive streamer discharge in air were accurately analyzed with 1 ns and 100 µm scales [132]. Recently, photoionization rates have been calculated using a three-exponential Helmholtz model and neural networks [135].

In direct-current-driven positive streamer coronas, the behaviors of self-pulsing discharges were analyzed using EFISH. The transient coronas were bridged between the electrodes at a frequency of ∼3 kHz and superimposed with glow coronas [136]. The nanosecond plasma streamers between the two electrodes under a pressure of 200 mbar of pure nitrogen operated with pulse durations of 200 and 250 ns and repetition rate of 1 kHz were measured using EFISH and coherent anti-Stokes Raman scattering (CARS). The CARS spectra could be analyzed with a vibrational excited distribution, and EFISH measured the maximal electric fields up to 81 Td [137]. Electric-field-induced CARS was demonstrated for measurements of electric fields in H₂ and N₂ containing discharges at pressures higher than atmospheric pressure [138,139]. The electric field was measured in a nanosecond plasma jet with helium and nitrogen flows using EFISH [140]. The speed jet propagates at the edge of the ionization waves. The spatiotemporal resolution was 1 ps and 50 µm scales.

Another method for determining electric fields includes the laser-induced fluorescence (LIF) technique [141], which can detect fluorescence dips caused by the Rydberg states of rare gases in an electric field. Depending on the selection of the Rydberg level of xenon, they demonstrated a measurement for electric fields of ∼100 V/cm [142].

Optical techniques are mainly used for plasma diagnostics at atmospheric pressure. Recently, significant progress has been made in highly sophisticated electric field measurements, especially for the EFISH method and high-speed imaging. Attention should be paid to their development.

Measurements of chemical properties of plasma

In plasma, electrons collide with the gases in the background. Thus, rich chemistry, such as excitation, dissociation, attachment, and ionization, is provided. For instance, differences in properties, such as electron density, oxygen atom density, and gas temperature, between the two plasma sources have been comprehensively examined using a combination of several methods [143]. In the effluent region of the plasma source, the chemical species were measured using VUV absorption spectroscopy, laser-induced fluorescence spectroscopy, and optical emission spectroscopy [144] (Figure 9). VUV absorption spectroscopy revealed the spatial distribution of nitrogen and oxygen atoms [145,146]. Under open-air conditions, various reactive oxygen and nitrogen species are generated by the gas-phase reactions, and the composition of the reactive species depends on the distance from the exit of the plasma source [144].

Figure 9. Absolute densities of O (³P) atom and N (⁴S°) atom generated by an Ar plasma source [144] (Modified from J Phys D 50, 195202 (2017)).

Nastuta and Gerling reported chemical concentrations of ozone and nitrogen oxides in the DBD jet flowing argon or helium and these concentrations depended on the strengths of the developed electric fields [147].

Excited nitrogen molecules were measured by tunable laser absorption spectroscopy (TDLAS) in the nanosecond DBD at pressures up to 5 kPa (400 Torr) of mixtures of H₂ and N₂, O₂ and N₂, and NO and N₂ [148].

Metastables of helium were measured at absolute concentrations and radial profiles via laser absorption spectroscopy at 1083 nm for helium (He ³S₁), in a condition of 8 ns rise and fall time with a repetition rate of 20 kHz. The metastable density at 2 × 10 ¹³cm⁻³ was affected by the oxygen byproducts, such as O atom, O₂ ¹Δ, ozone, and positive and negative ions [149]. The excited states of the nitrogen molecules were measured [150].

Through numerical simulations of the oxygen-mixed argon plasma jet, it was concluded that oxygen increased the scattering of electrons toward the targets and induced an increase in reactive oxygen species (ROS), and dissociative electron attachment (DEA) to water molecules is a major reaction of electron loss [151].

Measurements of physical behaviors of liquids contacted with plasma

Complex flow dynamics were observed in the liquid when the plasma streamer touched the liquid surface. First, surface discharges on the distilled water surface spread from the pin electrode. This phenomenon has been explained by the ionic winds creating a convection flow from the plasma-touching location to the outer surface. This generates downward and upward directions in the outer and inner portions of the liquid vessel, respectively [152]. It has also been observed that laminar and circulating flows occur in the liquid from the plasma-touching point into the bulk liquid [153]. Reactive oxygen and nitrogen species (RONS), generated at the plasma-touching point, move through the flows by advection and diffusion. The moving speeds of RONS inside the bulk liquid portions were analyzed via ultraviolet absorption spectroscopy. Furthermore, acidification and sparging of dissolved gases occur at the plasma touching point [154]. The exploitation of plasma-induced turbulence and electrohydrodynamic (EHD) forces on liquid impingement and the resulting convective flow has been initiated [155] (Figure 10). Liquid particle image velocimetry (PIV) (Figure 11), liquid pH imaging using bromothymol blue (BTB) agents (Figure 12), and gas flow patterns via Schlieren imaging were simultaneously performed. When the plasma touches the water surface, a cavity is formed and its dimensions are observed through imaging. The cavity depth and width increased proportionally with the applied voltage [155]. Currently, a consensus on cavity dimensions determined by a balance between the EHD force and buoyancy of the liquid has been reached [156]. Considering the pH and liquid temperature, the plasma cases exhibited higher flow velocities and larger circulation flows, where the liquid flow velocities were of the order of millimeters per second (mm s⁻¹) within the circulation flow and less than 1 mm s⁻¹ within the secondary flows.

Figure 10. (a) Schematic and (b) image of the atmospheric pressure jet (APPJ) (plasma plume was visible during operation) [155] (Reprinted from J Phys D 52, 075203 (2019)).

Figure 11. APPJ strikes at center top and streamline images 2 min into treatment at 5 mm nozzle elevation for (a) 2 slm helium, (b) 2 slm plasma at 8 kVpp, (c) 4 slm helium, and (d) 4 slm plasma at 8 kVpp (plasma discharge increased the liquid velocity and size of the circulation flow) [155] (Reprinted from J Phys D 52, 075203 (2019)).

Figure 12. Streamlines of pH changes measured using a BTB solution [after treatment over 5 min at 8 kVpp and 2 slm, acidity appears at the impingement point (t = 10 s), and acidified liquid is largely contained within the circulation flow; in the absence of jet-driven circulation flow, acidic liquid diffuses throughout the vessel after an initial downward flow caused by liquid inertia (t = 6 min)] [155] (Reprinted from J Phys D 52, 075203 (2019)).

The physical impact of plasma irradiation on bulk liquid flows depends on the jet impact force and additional shear across the liquid surface. This implies that liquid mixing is critical, and circulating flows also generate complex phenomena in relation to the jet velocity, adherence, liquid convection, and vessel shape.

Measurements of chemical properties in liquids contacted with plasma

In the gaseous reactions of plasma in ambient air, OH radicals are formed after the dissociation of water molecules. Furthermore, hydrogen peroxide (H₂O₂) and nitrous acid (HNO₂) are generated and dissolved in water. In addition, peroxynitrite (ONOO⁻) and peroxynitrate (O₂NOO⁻) are produced [157]. To understand the reaction mechanisms of RONS generation during plasma irradiation, free radicals in an aqueous solution were analyzed using the electron spin resonance (ESR) technique with the spin trapping method [158]. Figure 13 shows the schematic of the reaction model. The generated OH radicals in water can oxidize the organic constituents in the solution [158].

Figure 13. Schematic of reaction pathways in the plasma plume, gas phase, surface, and the bulk liquid [feed (blue), precursor (black), and antitumor (red) species] [158] (Reprinted from J Phys D 50, 155202 (2017)).

Further examination of free radical generation was conducted using an alcohol-water mixture solution with 3,5-dibromo-4-nitrosobenzene sulfonate (DBNBS) as the spin trap agent [159]. Notably, the plasma cases differ from those of the pyrolysis reaction. The spin adducts of the ·CH-and ·CH₂-type radicals were mainly formed by H-abstraction by ·OH radicals. The formation of the H-abstraction radicals by ·OH radicals formed abundantly in aqueous solutions, is predominant in the plasma chemistry [159] (Figure 14). Recently, it was observed that generation of H₂O₂ was similarly inhibited in hydroxy acid-water mixtures, and it is evident that ·OH radicals in aqueous solution oxidize the organics in solutions [160].

Figure 14. Reaction schemes of (a) methanol, (b) ethanol, and (c) 1-propanol (chemical reactions suggested that hydrogen abstraction was indirectly dominant) [159] (Reprinted from J Phys D 51, 095202 (2018)).

It is evident from the aforementioned studies that plasma-activated organics have attracted considerable attention for the clarification of aqueous reactions induced by the plasma irradiation of liquids. When a mixed solution of water and lactic acid (hydroxy acid) was irradiated with plasma, acetyl, and pyruvic acids were detected in the treated solution through nuclear magnetic resonance (NMR) analysis [161]. Meanwhile, using the direct infusion electrospray ionization tandem time-of-flight mass spectrometry (ESI-TOF-MS/MS), glyoxylic acid, and 2,3-dimethyltartrate were detected [161]. In addition, when a mixture of argon, nitrogen, and oxygen gases was used for plasma generation, the plasma-irradiated lactic acid solutions contained products with methyl amine groups [162]. The plasma-induced aqueous reactions include dehydration, esterification, hydrolysis, and dimerization of the original organic molecules [163]. Notably, rich chemistry for the production of substances derived from the initially added lactic acid was found [164].

Future beyond measurements of low temperature plasma-driven phenomena

It is well-known that solar wind can induce electron collision reactions in the Earth’s atmosphere (Figure 15). Similar to the Sun–Earth relationship, interactions between the plasma and liquid promote artificially heterogeneous reactions of energetically excited particles. Subsequently, the plasma-activated reaction fields relax with secondary and further reactions proceeding kinetically. We believe that the gaseous concentration of oxygen determines the ·OH-mediated aqueous chemistry. Thus, the plasma-activated organics anticipate the antitumor effects (Figure 15). These processes are complex and continue to be systematically studied.

Figure 15. (Left) Schematic of the interaction of solar wind with earth’s atmosphere, (right) concept model of plasma-activated organics generated in culture medium fluids; rich RONS and hydroxyl radicals mediate the aqueous chemistry; PAM produced under hyperoxia in ambient conditions has potent cytotoxicity anticipated by plasma-activated organics.

In summary, the hierarchy from gaseous physical plasma to aqueous chemistry is essential for providing rich free-radical chemistry (Figure 16). Plasma-induced electron collisions ensure kinetically driven chemical derivatives in association with air and water. These reactions occur sequentially and hierarchically via physical plasma dynamics, gaseous chemical kinetics, plasma-liquid interfaces, aqueous chemistry, plasma biological interfaces, and biological responses. Herein, there are two schemes: direct irradiation with plasma-biological interactions and indirect irradiation with plasma–liquid interactions. In the latter case, the biochemical and biological effects occur through biochemically reactive species in plasma-treated liquids. Currently, we are focusing on the plasma-activated organics derived from sugars, lipids, amino acids, and proteins. Understanding the mechanisms of cell death has important implications for the development of plasma cancer therapies.

Figure 16. Development timeline of representative diagnostics for plasma–liquid and plasma-biological interactions. LTS: laser Thomson scattering for electron density and temperature measurements; ESR: electron spin resonance. Jia 2011 [165], Ishikawa 2012 [166], Takeda 2013 [167], Jia 2014 [168], Takeda 2017 [144], Takeda 2019 [143], Brubaker 2017 [154], Brubaker 2019 [155], Tanaka [161], Liu [163], Uchiyama [159], Kurake 2016, 2017 [158,169], Kurake 2019 [170], and Ishikawa 2020 [171].

Chronological development of plasma cancer therapy

Brief overview of development timeline of plasma cancer therapy

In 2002, Stoffels et al. demonstrated glow discharge using a 0.1 mm tip metal electrode at atmospheric pressure, called as the “plasma needle.” They proposed its usage in surgical and dental instruments, and cancer therapy [101].

In 2007, floating electrode dielectric barrier discharges (FE-DBDs) were used in this field, and the potential medical applications of low-temperature atmospheric pressure plasma were found to be the sterilization of the affected areas of trauma and the promotion of wound healing [102].

In the 2010s, microplaster, kINPen, and non-equilibrium atmospheric pressure plasma (NEAPP) sources were developed and were used to test cancer cell treatments. These devices have been used for both clinical and biological treatments. Briefly, there are two trends. One is the preclinical tests in dermatology and dentistry as surface treatments based on plasma-biological interactions. The other is a type of drug- and solution-based study using plasma-treated liquids, which results in plasma–liquid interactions. Historically, oxidative species, such as ozone, H₂O₂, superoxides, and nitrogen oxides, such as nitrous and nitric oxide anions (NO₂⁻ and NO₃⁻), peroxyl nitrogen oxide anions (ONOO⁻ and O₂NOO⁻) have received attention. Subsequently, studies have focused on the derivatives of liquid components, called plasma-activated organics. The mechanisms underlying cell death have been elucidated, and are roughly categorized as apoptosis, immunological cell death, autophagy, and ferroptosis, as described later.

Electrocautery scalpels are used as medical instruments, and argon plasma coagulation devices are used to stop bleeding. Although these devices use high-temperature plasma and work differently, the development of treatment devices using low-temperature plasma and medical technology using such devices are also in progress [172]. Important progress in this field has been made by changing the regime from high-temperature plasma devices to low-temperature devices.

As of 2022, certification of electrotherapy devices is in progress in the United States and Europe, and full-scale commercialization is expected in the near future. The safety and efficacy of electrothermal therapeutic instruments for cancer treatment have also been verified [173].

The development timeline of plasma cancer therapy is shown in Figure 17.

Figure 17. Development timeline of plasma cancer therapy. FE-DBD: floating electrode dielectric barrier discharge; NEAPP: non-equilibrium atmospheric pressure plasma; LF-jet: low frequency operated plasma jet; PAM: plasma activated medium; PAL: plasma-activated ringer’s lactated solution; EBM: evidence-based medicine; ICPM: international conference on plasma medicine; IWPCT: international workshop on plasma cancer therapy. Akiyama [174], Stoffels [101], Kieft [175], Fridman [176], Kalghatgi [102], Isbary [103], von Woedtke [104], Kaushik [177], Vandamme [178], Lukes [179], Miller [180], Metelmann [181], Sato [182], Iwasaki [12], Iseki [85], Ikehara [106], Tanaka 2013 [86,87], Kurake 2016 [169], Tanaka 2017 [183], Kurake 2019 [170], and Jiang [184].

Progress of preclinical studies on plasma cancer therapy

In Japan, Sato et al. reported that H₂O₂ is a key inactivation factor in HeLa cell viability by a plasma flow [182]. In 2013, Tanaka et al. found that plasma-irradiated cell culture media selectively killed cancer cells [86], In 2014, Kikkawa et al. published the first report on the tumor-shrinking effects of a drug-resistant cancer cell line in a mouse model [87]. Kurake et al. reported the synergistic effect of H₂O₂ and NO₂⁻ on the selective death of cancer cells [169]. In 2016, the selective killing of cancer cells was also found in tumors after intravenous injection of plasma-treated Ringer’s solution [183].

Microarray analysis of the gene expression profiles of gliomas cultured in PAM and plasma-irradiated Ringer’s solution (PAL) revealed characteristic differences in the expression of antioxidant genes and signaling molecules for survival and growth [185] (Figure 18). The reactive oxygen species (ROS) in cells cultured with PAM and PAL were observed after staining the cells using DCF (2′,7′-dichlorodihydrofluorescein diacetate) as fluorescent probes. In PAM-treated cells, the fluorescence disappeared after treatment with NAC (N-acetylcysteine) as a ROS scavenger. Thus, the ROS stress caused by the PAM in cells can be eliminated by treatment with NAC. ROS stress in PAL-treated cells was low. Expression levels of the antioxidant genes catalase (CAT), superoxide dismutase (SOD2), and glutathione peroxidase (GPX1) were not elevated in cells treated with PAM or PAL. In contrast, the expression of cyclin-dependent kinase inhibitor 1A (CDKN1A) and other genes involved in growth arrest and DNA-damage-inducible protein (GADD45α) signaling pathway was upregulated only in PAM-treated cells. CDKN1A, a protein-coding gene, is postulated to induce apoptosis by arresting the cell cycle. Proliferation signals are activated in cancer cells, one of which is the regulation of gene expression by the activation of various transcription factors through the phosphorylation of mitogen-activated protein kinases (MAPK). The inhibition of this pathway induces cell death. In contrast, oxidative stress factors are not upregulated in PAL-treated cells. We speculate that the downregulation of activating kinases (c-Fos and c-Jun) downstream of the MAPK signaling pathway leads to cancer cell death by downregulating the expression of activators in the cell proliferation signaling cascade. Ringer’s lactate solution and culture medium share the same salts (NaCl, KCl, and CaCl₂), and the main differences are that the Ringer’s lactate solution includes L-sodium lactate, while the culture medium does not include L-sodium lactate, and the culture medium includes more than 30 components, while the Ringer’s lactate solution contains only four components. Therefore, we believe that the PAM-induced cell death is more complex than the PAL-induced cell death.

Figure 18. Intracellular molecular mechanisms to explain the differences between PAM- and PAL-treated glioblastoma cells; models of intracellular molecular mechanisms of cell death in (a) PAM-treated and (b) PAL-treated glioblastoma cells, based on microarray and qRT-PCR [185] (Reprinted from Sci Rep 9, 13657 (2019)).

The metabolite profiles of brain tumor cells treated with the PAM and PAL were examined [170]. The metabolism of the glycolytic system, which produces pyruvate and ATP from glucose, known as the Warburg effect, is enhanced in cancer cells, and the metabolites are retained in the glycerol 3-phosphate (G3P) step of the glycolytic system (Figure 19). This is interpreted as the inhibition of the step, in which the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) catalyzes phosphorylation to pyruvate via bisphosphoglycerate (BPG), and a decrease in nicotinamide adenine dinucleotide (NAD⁺), a coenzyme of GAPDH. NAD⁺ is synthesized from nicotinamide mononucleotide (NMN) by nicotinamide phosphoribosyltransferase (NAMPT). Nagaya et al. showed that the inhibition of NAMPT using the inhibitor FK866 markedly enhanced the PAM-induced cell death in human breast cancer MDA-MB-231 cells [186]. In the PAM-treated cells, decreased NAD⁺ levels induced cell death.

Figure 19. Model of NAD-enzymatic deactivation of PAM treated cells. NMNAT enzyme deactivated and accumulated PRPP. ATP and NAD⁺ synthesis downregulated and a key GAPDH enzyme deactivated. Metabolism in down-flow from G3P to BPG in glycolysis was deactivated with intermediate accumulation [170] NAM: nicotinamide; NAMPT: nicotinamide phosphoribosyltransferase; NMN: nicotinamide mononucleotide; NMNAT: nicotinamide mononucleotide adenylyltransferease; G3P: glyceraldehyde 3-phosphate; BGP: 1,3-bisphosphoglycerate. (Reprinted from Arch Biochem Biophys 662, 83 (2019).).

The metabolite profile of the PAL-treated cells showed a high ratio of reduced glutathione compared to oxidative stress, a characteristic of the absence of oxidative stress in cells [171] (Figure 20). Cells become hypotrophic after treatment with the Ringer’s solution and the supply of nutrients, including glucose, is depleted, resulting in enhanced anabolism. Mitochondria, which are responsible for cellular energy production, produce ATP via the oxidative phosphorylation of the tricarboxylic acid (TCA) circuit. Alpha-ketoglutarate (KG), also found in the TCA pathway, is synthesized from glutamate by glutamate dehydrogenase (GDH), using NAD⁺ as a cofactor. This is a reaction, in which an amino group is transferred from an amino acid to synthesize a keto acid and is considered similar to the reactions between the glutamic acid and oxaloacetic acid and between alanine and pyruvic acid. Amino acid metabolism, including the hydrolysis of aspartic acid from asparagine, was also significantly altered in the PAL-treated cells.

Figure 20. Model summarizing metabolomic changes in PAL-treated cells in comparison with the previous study on PAM [171] (Reprinted from Arch Biochem Biophys 688, 108414 (2020)).

Cancer cells require substance synthesis for self-renewal because of their active cell proliferation, and adipogenesis from acetyl-CoA is also enhanced. PAL treatment induces anabolic intracellular metabolism, and despite the low oxidative stress, it inhibits anabolic amino acid and fatty acid synthases, leading to cell death. A major difference in the PAM is that cell death is induced by oxidative stress and reduced energy production, whereas intracellular catabolism is still in progress.

In terms of intracellular energy production, glucose is metabolized to pyruvate by the glycolytic system, and pyruvate is converted to acetyl-CoA and citric acid to produce ATP via the mitochondrial TCA circuit. In oxidative phosphorylation, an enzyme system called the electron transfer system reduces oxygen to water instead of using oxygen to oxidize coenzymes; ATP is synthesized by phosphorylation from the concentration gradient of hydrogen ions inside and outside the mitochondrial membrane. This process allowed us to evaluate the main energy metabolic pathways of the cell based on the extracellular proton concentration and oxygen consumption. The extracellular acidification rate (ECAR) and oxygen consumption rate (OCR) of the PAL-treated cells can be measured using an extracellular flux analyzer system that performs this evaluation. At the time of measurement, glucose was introduced to activate the glycolytic system, followed by the introduction of oligomycin (OM), which inhibited the mitochondrial ATP synthesis, thus maximizing the metabolic capacity of the glycolytic system. The subsequent introduction of 2-deoxyglucose (2-DG) stops glucose metabolism, and the metabolic capacity of the glycolytic system can be evaluated. In addition, by introducing the mitochondrial deconjugating agent p-trifluoromethoxy-phenylhydrazone (FCCP) and the electron transfer system inhibitor antimycin (R/AA), the oxygen respiratory capacity of the mitochondria can be evaluated. The results of this evaluation in the PAL-treated cells showed that the glucose metabolism and TCA cycle were suppressed [187] (Figure 21). The mitochondrial membrane potential decreased in the PAM-treated cells, and no decrease in membrane potential was observed in the PAL-treated cells; however, ECAR was unchanged in the FCCP and R/AA treatments. These findings suggest that the glycolytic system and TCA cycle, but not the electron transport system, are inhibited.

Figure 21. Inhibition of glycolysis and tricarboxylic acid (TCA) cycle in HeLa cells by plasma-activated ringer’s lactate solution (PAL); extracellular H₂O₂ induces intracellular H₂O₂ by penetrating through aquaporins, subsequently inducing apoptosis in HeLa cells; non-H₂O₂ PAL components in PAL (PAL2w) induce intracellular ROS (but not H₂O₂), which are responsible for impairment of mitochondrial respiratory system, glycolysis, and TCA cycle, ultimately inducing non-apoptotic cell death in HeLa cells (arrowheads represent stimulatory relationships and blunt arrowheads represent inhibitory relationships) [187]. (Reprinted from Plasma Process Polym 18, 2100056 (2021)).

Yoshikawa et al. treated endometrial cancer cells with PAM, which inactivates the mTOR, a signaling pathway that regulates autophagy during nutrient depletion. In addition, inhibition of the autophagy inhibitory factor (MHY1485) induces autophagy-induced cell death. The accumulation of cells in the G2/M phase of the cell cycle also suggests autophagic cell death [188]. Nakamura et al. reported the cell-killing effects of PAL produced by a commercially available novel PAL manufacturing device, in which plasma was generated by introducing pure nitrogen or pure oxygen rather than air [133]. When ovarian cancer metastasizes to the peritoneum, scattered microscopic foci are formed. Using a mouse model, we reported the possibility of treating these cancer foci by intraperitoneally circulating PAL and their disappearance [189].

Toyokuni et al. noted that divalent iron is abundant in cancer cells and is a source of ROS generation. Therefore, they explained the mechanism, by which the ratio of sulfur and iron, which are abundant in enzymes with antioxidant activity, causes selective cell death in cancer cells. Autophagy helps the intracellular degradation of damaged lysosomes as an inflammatory response; inducible NO synthase (iNOS) is activated through NF-κB, a transcription factor involved in inflammation, to produce NO, which induces peroxynitrite production in the lysosomes in the presence of superoxide. We reported that as this process progresses and reaches a late stage, membrane permeability of the lipid peroxide-derived lysosomes increases, and ROS leaks out and becomes excessive in the cell, leading to cell injury in the presence of divalent iron [184] (Figure 22). This form of cell death, derived from divalent iron and ROS, has recently been referred to as ferroptosis [190]. Okazaki et al. reported that OH and O₂- radicals exert oxidative stress on biomolecules by spin-trap ESR for the generation of biological radicals by low-temperature plasma irradiation when dithiothreitol, reduced and oxidized glutathione, or dehydroascorbate were added as antioxidants [191,192]. From these results and other literature on the nitrosylation of iron-sulfur complexes (S-nitrosylation), it can be inferred that an abnormal redox balance involving iron, sulfur, and NO induces cell death specific to cancer cells.

Figure 22. Model of PAL-induced ferroptosis in malignant mesothelioma (MM) cells, with exogenous NO as a component of PAL; initially, NO activates transcription factor NF-κB in MM cells, upregulating the downstream iNOS; then, NO accumulates in lysosomes, starting autophagic process; eventually, lysosomal lipid peroxidation is exceeded over the threshold and leads to lysosomal membrane permeabilization (LMP); this entire process eventually leads to ferroptosis [184] (Reprinted from Redox Biol 43, 101989 (2021)).

Current global status of plasma cancer therapy

Since 2012, the Max Plank Institute in Germany has taken the lead in the field of wound healing. Today, in Greifswald, at the INP Institute, the group of Bekeschus and Wende actively promotes plasma biotechnology research and has developed a plasma source named kINPen, which is commercially available for medical, dental, and veterinary use. They reported that the expression of immunogenic cell death (ICD) markers was enhanced and macrophages and T cells were activated when a plasma-irradiated solution was applied to 3D spheroids of mouse colon cancer cells [193]. ICD was not observed with H₂O₂ alone and was unique to plasma-irradiated solutions [194]. Cancer recurrence sometimes leads to microbial infection, immunodeficiency, and stromal cell proliferation, resulting in scarring [195]. When the pH of the liquid phase decreases during plasma irradiation, interferon gamma (INFγ) and interleukins (IL2, IL6, IL10, and IL17F) decrease, based on the evaluation of mouse models [196]. This review focuses on the roles of immunity. Plasma treatment regulates immunological activation. Others have reported that there is no correlation between the expression of aquaporins and redox-related enzymes, such as NADPH oxidase, on the plasma membrane and that the action of plasma is strongly correlated with cancer metabolism [197]. Wende et al. comprehensively examined the molecular modifications of plasma-irradiated amino acids and highlighted the importance of cysteine oxidation [198]. Based on the results of plasma irradiation of zebrafish, an attempt was made to evaluate their safety [199]. Moreover, Metalmann is working on palliative treatment in end-of-life care at hospitals as a fact- or evidence-based treatment (EBM) [11].

In the US, Keidar et al. reported that plasma irradiation combined with an anticancer drug, temozolomide (TMZ), for glioblastoma [200] and with FOLFIRINOX for cholangiocarcinoma promoted cytocidal effects [201]. The combination of iron oxide-based magnetic nanoparticles and plasma irradiation suppressed the activity of proteins downstream of epidermal growth factor receptor (EGFR) signaling (phosphorylation of ERK and Akt), induced apoptosis in A549 lung cancer cells, and suppressed xenograft tumor growth [202].

Focusing on the physical effects of plasma, Yan et al. investigated the effects of plasma treatment on glioblastoma cells. For the first time, they demonstrated that physical factors in plasma could reinforce the positive selectivity of plasma-treated astrocytes [203]. Other studies have reported the application of magnetic fields to sensitize the melanoma cell line B16 to the cytotoxicity of ROS [204] and compared the physical and chemical effects on cell death using four typical cancer cell lines [205]. When a three-dimensional tissue model made of collagen was irradiated with a He plasma jet, an electric field of ∼300–30 kV/m was generated at the cell membrane surface potential [206]. This is considerably lower than an electric field of 1 MV/m or higher, which perforates the cell membrane.

In Europe, research teams working on the treatment of bone and soft tissues, are focusing on plasma application for cancer treatment targeting osteosarcoma. Canal et al. have recently been active in plasma medicine within the scope of osteosarcoma therapy [207]. They reported significantly higher production of reactive species in the hydrogel (2% gelatin) polymer solution than that in water after plasma treatment. Furthermore, while the viability of osteosarcoma SaOS-2 cells decreased upon exposure to plasma-treated gelatin, the viability of healthy cells was preserved (∼90%), establishing the selectivity of the plasma-treated gelatin for cancer cells [208]. When pyruvic acid was included in the plasma-irradiated solution, the amount of H₂O₂ and nitrite (NO₂⁻) generated selectively killed cancer cells [209]. They also reported the effect of plasma-irradiated Ringer’s solution treatment on osteosarcoma by creating scaffolds that serve as bone models using collagen and hydroxyapatite, and constructing 3D tissue models [210].

Around 2014, ICD began to attract attention, and the first notable publication on the plasma activation of immune responses was by Miller et al. If cell death is caused by a secondary effect of immune activation, the plasma irradiation effect will be widespread, resulting in ICD attracting considerable attention [180,211–213].

Eloisa et al. treated metastatic melanoma and pancreatic cancer cells with a plasma-irradiated culture medium, resulting in the suppression of cell proliferation, which promoted a decrease in ATP levels and an increase in calreticulin, activating ICD [214]. T lymphoblasts were cultured in a plasma-irradiated culture medium under hypoxic conditions, resulting in selective lymphocyte activation [215].

A group at Old Dominion University demonstrated 3D tumor spheroids in an experimental system [216]. They reported the combined effects of plasma and anticancer agents (cisplatin, methotrexate, Adriamycin, and paclitaxel) on 3D bladder tumor spheroids [217], hyperthermia [218], and electrochemical treatment [219].

A collaborative study by the Jiangnan University of China and Ostrikov group of Australia reported quantitative assessments of the parameters of PAM preparations that determined the anti-tumor efficacy of PAM-cultured cells [220]. Endogenous apoptosis has traditionally been shown to decrease the mitochondrial membrane potential, suppress Bcl-2 expression, increase Bax expression, and induce apoptosis [221]. Plasma treatment protects neural SH-SY5Y cells from oxygen-induced apoptosis and glucose deprivation owing to the intracellular NO production [222]. Other authors have reviewed the results of plasma membrane transport by plasma, known as plasma poration [223,224].

Recently, the enhancement of apoptosis due to the effect of PAM alone and in combination with carboplatin in chemotherapy-resistant ovarian cancer cells has been reported [225].

Various theories have been proposed for the mechanism of selective cancer cell death. Baur et al. speculated that catalase on the cell membrane surface of cancer cells is inactive and that H₂O₂ and NO₂⁻ act synergistically to generate secondary singlet oxygen when the cells are cultured in plasma-irradiated culture medium. This singlet oxygen affects the redox enzymes (NOX and NOS) in neighboring cells and induces apoptosis in the surrounding cells [226].

Conway et al. have demonstrated a caspase-independent mechanism of cell death in p53-mutated glioblastoma multiforme cells exposed to plasma. Lysosomal markers are expressed, lysosomes accumulate, and caspase-independent cell death occurs [227]. In 2014, caspase-independent cell death was reported by Adachi et al. and the pursuit of identifying the plasma-specific cell death mechanism is now underway [228].

In a separate experiment with A549 (human lung cancer) and MDA-MB-23 (human mammary cancer) cells, it was reported that the expression of Sestrin2 (Sesn2) was induced, JNK was activated, and LC3 that binds to the autophagic membrane was formed, leading to the induction of autophagy. These results indicate that the JNK/Sesn2 pathway is involved in cell death induced by non-thermal plasma [229].

Bourdens et al. reported that plasma irradiation of adipose mesenchymal stromal cells differentiated in the stroma increased p53/p21 expression, inhibited cell proliferation, and promoted p16 expression. This is a cellular senescence phenomenon, and the expression of-galactosidase, a marker molecule for this phenomenon, and the secretion of inflammatory cytokines have also been observed [230].

Notably, it has been reported that plasma irradiation integrated with microneedles for transdermal delivery combined with a checkpoint inhibitor (PDL1) in immune checkpoint blockade therapy increases treatment efficacy and inhibits the growth of both primary tumors and distant tumors, prolonging the survival of tumor-bearing mice [231]. The killing effect of plasma on drug (temozolomide; TMZ)-resistant cell lines was reported by Morfill’s group in 2013 [232]. Positive results were obtained for many patterns using a combination strategy; however, more research is needed to monitor the intracellular conditions and examine the mechanisms of action.

In a mouse model of drug (gefitinib)-resistant lung cancer, Choi et al. reported that oral administration of plasma-irradiated water induced immune activation and debilitates cachexia due to malnutrition [233]. Anticancer drug-resistant A375 melanoma cells have been reported to be killed synergistically by He + O₂ plasma irradiation and doxorubicin [234].

Lee et al. found that melanin production in normal human skin melanocytes was enhanced by plasma-irradiated culture medium and speculated that this was owing to the increased expression of tyrosinase, a synthetic enzyme in the skin [235]. Treatment of mast cells with plasma-irradiated culture medium alleviates atopic dermatitis-like symptoms and suppresses IL6, TNFα, and NFκβ. On the contrary, differentiation of Th2 cells, a type of helper T cell, is induced, and skin inflammatory cytokines and chemokines are weakened [236]. Liu et al. reported that plasma-activated medium exerts anti-invasion and anti-metastatic effects on B16 melanoma cells, mainly by suppressing the-catenin pathway, which consequently decreases the expression of MMP-9, MMP-2, and CD44 via H₂O₂ [237]. Inflammation is alleviated by plasma-specific mechanisms.

Jezeh et al. compared the effects of direct plasma irradiation and plasma-activated medium in HeLa and MDA-MB-231 cells and found that the expression levels of apoptotic proteins, caspase-3, and caspase-8, were higher with direct plasma treatment than with PAM treatment [238]. A morphological risk assessment of cold atmospheric plasma-based therapy using bone marrow mesenchymal stem cells in the proximity of the treatment zone was reported by Hajizadeh et al. They observed cell and nuclear morphology, examined the relationship between the expression ratio of Bax and Bcl2, and showed cell damage in high-power plasma sources and high-dose plasma-activated water. Moreover, it was concluded that the probability of cancer initiation is a critical issue that should be investigated in all plasma-based cancer therapies, even in optimized indirect methods [239]. All these experiments are relevant to the mechanism of cell death [240]. However, there are uncategorized mechanisms. Abnormal cell death with “spongy” structures has also been reported [241]. Caspase-independent apoptosis is also induced by the release of cytochrome c from mitochondria. These mechanisms are referred to as endogenous apoptosis, and there is an exogenous pathway involving the receptor (TNF). Further investigation of plasma-induced cell death mechanisms is required.

Plasma irradiation inhibits mitosis-related chromosome division and DNA replication. Throughout the cell cycle, S- and M-phase cancer cells are more susceptible to plasma irradiation than the G₁- and G₂-phase cells [242]. DNA replication does not occur in human retinoblastomas irradiated directly or indirectly with plasma. Nevertheless, there was no increase in glutathione levels, which is a result of protection against oxidative stress, and the number of cells that remained in the S and G2/M phases increased without DNA strand breaks. Here, they indicated that the reactive species increase comes in two timely distant waves, the first originating from the plasma itself with secondary solubilization and passive diffusion, and the second derived from the mitochondrion. Based on these results, they considered that the combination with an anticancer drug (carboplatin) was effective and reported an actual increase in cell death [243].

Hara et al. suppressed nicotinamide phosphoribosyltransferase (NAMPT) related to glucose metabolism in plasma-activated cell culture medium (PAM) with a NAMPT inhibitor, FK866, resulting in decreased NAD and ATP levels owing to the mitochondrial damage. Intracellular reduced glutathione also decreased, suggesting the beneficial effects of the combination of FK866 and PAM [186].

Jawaid et al. reported size-dependent effects of gold nanoparticles and showed that the combination of 2 nm nanoparticles and low-temperature plasma in human lymphoma U937 cells induces enhanced apoptosis due to the increase in the oxidative level and loss of antioxidant capacity [244].

There are several methods for treating drug-resistant cancer cells strains that relapse with a poor prognosis. Current chemotherapies are problematic because of their numerous side effects. Developing a method with fewer side effects using plasma medicine is a promising approach. The use of liposomes and nanoparticles to concentrate the therapeutic effect on the affected area is promising [245,246]. With many reports of clinical applications, it is expected to be established as a treatment method different from conventional medicine.

In the mouse model, rapid abscopal effects were observed, suggesting innate immune response activation. The paper reported on the B16-F10 tumor growth at a non-irradiated site on a mouse leg after plasma was applied to the tumor in the other leg. Tumor growth suppression at non-irradiated remote sites was observed the day after the plasma irradiation [247].

In 2010, research was conducted on the use of cold plasma in cancer therapy. Primarily, these in vitro studies have been limited to skin cells and simple cellular responses to cold plasma treatment [248–250]. Additionally, preliminary reports on the in vivo antitumor effects of plasma using U87 glioma xenografts have been published [178]. Subsequently, one of the unique features of low-temperature plasma in selectively eradicating cancer cells without damaging normal cells was reported [251].

Recently, comprehensive preclinical studies on cold atmospheric-pressure plasma in cancer treatment have primarily focused on in vivo studies over the past decade [252]. In addition, human clinical trials on the plasma treatments for head and neck cancer have been systematically reviewed [253]. Herein, we report five clinical consequences of plasma treatment. A total of 159 patients were included in the different study setups. Fifty-three patients were diagnosed with head and neck cancers and six with oropharyngeal squamous cell carcinoma; in these cases, kINPen MED using Ar plasma was used directly on the ulceration, vertical to the naturally moist tissue surface [181,254–256]. In a study of 100 cases of laryngeal carcinoma, the Unitec low-temperature plasma operation system was employed to directly ablate the tumor up to 3–5 mm away from the edge of the lesion [257]. Outcomes varied among the studies addressing the effects of plasma on the contamination of infected ulcerations, side effects, pain, tumor surface changes, and tumor growth. Although a few patients experienced mild-to-moderate side effects, a significant reduction in the fetid odor, contamination, and pain of lesions was observed.

Perspectives

Cancer therapy is a multidisciplinary treatment that combines surgery, chemotherapy, radiotherapy, and immunotherapy. Despite the elucidation of cancer mechanisms by comprehensive genomic and epigenomic analyses and the development of molecular therapies, drug resistance, and severe side effects present challenges to the development of new therapies [258]. Herein, we review the research trends in plasma-based cancer therapy up to 2022 (Table 1). Direct plasma treatments are physically limited to the reached depth and irradiated area. Drug resistance limits the effectiveness of pharmacological agents, and can lead to disease progression. Notably, plasma cancer therapy has been gaining interest in overcoming drug resistance in cancer to enhance the effectiveness of chemotherapy drugs [259]. Direct plasma treatment enhances the cancer-immunity cycle, as immune cells in both the tumor and tumor-draining lymph nodes appear to be more stimulated to perform their anti-cancer functions [260]. Thus, systemic plasma treatments are effective when combined with existing local physical modalities, such as radiotherapy and photodynamic therapy [261]. The immunostimulatory properties of the toxic actions of plasma-treated solutions are being tested by direct injection into the bulk tumor in a clinical setting and by lavage in patients suffering from peritoneal carcinomatosis as an adjuvant to standard chemotherapy [262]. Cells derived from solid tumors are less likely to undergo apoptosis and early apoptosis is rare. After plasma treatment, the mechanisms underlying the regulated cell death were diverse (Table 2). Cell types are diverse, and the modes of regulated cell death are yet to be discovered. Immunogenic cell death is of particular interest because local therapeutic effects affect the entire body.

Table 1. Brief summary of reports on direct and indirect plasma irradiations.

Direct plasma irradiation
kINPen med	ICD [193]
Helios (DBD), etc.	Physical factors [203,205] Magnetic factor [204]
mild plasma	Blood coagulation [106]
Combinations	Nanoparticles [202,244] Anticancer agents [217] Anticancer agents + PAM [225]
Preclinical studies	Palliative treatments [181] Laryngeal carcinoma [257]
Plasma pharmacy (indirect irradiation)
PAL (Ringer’s solutions): inorganics + hydroxy acid
	[183,185] Osteosarcoma [210]
PAM (culture medium): inorganics + glucose + amino acids + vitamins
	[86,87,169] Melanoma and pancreatic cancer cells [214,220] Human skin melanocytes [235] Inflammatory [236,237]
Plasma-treated gelatin	[208]

PAM: plasma activated medium; PAL: plasma activated ringer’s solutions; ICD: immunogenic cell death.

Table 2. Categories of typical mechanisms of regulated cell death [263].

Cell death mode	Related factors	Notes
Necroptosis	Necrosome (β-amyloid proteins), etc.
Pyroptosis	Membrane pore formation
Intrinsic apoptosis	Caspase-3; Permeability of the mitochondrial membrane
Extrinsic apoptosis	Caspase-8, caspase-3	Fas receptor signaling
Ferroptosis	GPx4	Fe^3+ to Fe^2+, ROS
Autophagy	Autophagosome; nucleus and cell membranes are healthy
Immunologic cell death	Calreticulin, etc.

GPx4: glutathione peroxidase 4; ROS: reactive oxygen species.

Plasma oncology has recently attracted considerable attention [264] and its final clinical application will be achieved after many preclinical studies [265]. The current state of the art is that the path to clinical applications has just opened, and it is hoped that further research in this field will be stimulated.

Notably, plasma has the potential to selectively kill cancer cells in contrast to the normal cells, not only through physical actions, such as the electric fields created by plasma but also through the direct and indirect actions of active chemical species in the living body. In 2014, a new phase of bio-applications of plasma was introduced in our previous research. Considering the previous situation, the action of plasma in sterilizing or inactivating living organisms is linearly linked to one parameter: the life or death of the organism. Concisely, the mainstream viewpoint has been “How can we make the plasma work?” (Figure 23(a)). As of 2022, there have been significant developments in the understanding of the principle of selective death of cancer cells in relation to normal cells. Through this development, the focus of research has shifted to the control of the mechanism of plasma action, meaning that the control of H₂O₂, NO₂⁻, and other organic compounds, short-lived reactive species, and reaction intermediates has become more important. The transport of these reactive species into cells is being considered, and new measurements of reactive species in the liquid phase are required to understand the state of cells in the field where they live. Understanding the formation, transport, and reactions of these reactive species is advancing simultaneously, and the most recent efforts to elucidate the mechanisms of action of plasma in living organisms have evolved from a linear relationship to the understanding and control of indirect effects, such as the selective synthesis of reactive species, their direct effects on cells, and understanding of intermediates and signaling cascades (Figure 23(b)). In addition, it is essential to develop plasma processing to understand and control the complex systems of living organisms in the future. The various active species of plasma, including the electric fields, ions, radicals, and light, are complex; however, the goal is to precisely control them to achieve the desired biological responses, such as selective cell death of cancer cells, through innovations in plasma processing technology (Figure 23(c)). Although this goal cannot be achieved without overcoming many obstacles, we hope that it will be realized toward the creation and development of a new academic field in plasma biology.

Figure 23. (a) Conventional model of plasma effects via linear cascade reaction responses, (b) present status of modeling of plasma-induced effects via wide-spreading synergisms on multiple reactions pathways, and (c) future goal of modeling via complicated reaction network with communication between clusters of bacteria, organelle, cells, etc. [14] (Reprinted from Jpn J Appl Phys 61, SA0805 (2022)).

Conclusions

In the state-of-the-art low-temperature plasma science, the history of plasma medical science and technologies for the generation and measurement of low-temperature plasma are outlined. Low-temperature plasma sources are useful for artificially promoting the effects of free radicals, even against environmental stresses. Thus, new therapeutic efficiency and efficacy can be expected from the use of low-temperature plasma, which can be evaluated using sophisticated quantitative measurements. Plasma cancer therapy is expected to be initiated soon.

Acknowledgments

The authors would like to thank professors Dr. Fumitaka Kikkawa, Dr. Keiji Imoto, and Dr. Akio Komori for encouraging this work; and Professor Dr. Masafumi Ito and the Hori-Ishikawa laboratory of Nagoya University for fruitful discussion and technical assistances.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This is supported in part by KAKENHI Grants-in-Aid (nos. 19H05462, 17H02805, 20H00142, and 21H04451) from JSPS and by the Center for Low Temperature Plasma Science, Nagoya University.