Producing nitric oxide by pulsed electrical discharge ...

RESEARCH ARTICLE PULMONARY HYPERTENSION

Producing nitric oxide by pulsed electrical discharge in air for portable inhalation therapy Binglan Yu,1 Stefan Muenster,1 Aron H. Blaesi,1 Donald B. Bloch,1,2 Warren M. Zapol1*

INTRODUCTION Administration of nitric oxide (NO) gas by inhalation produces localized relaxation of the pulmonary vasculature without dilating the systemic vasculature. NO mediates a broad scope of biological effects in mammals, including vasorelaxation, bronchodilation, inhibition of mitochondrial respiration, platelet and leukocyte activation, and regulation of smooth muscle proliferation. Inhaled NO was approved by the U.S. Food and Drug Administration in 1999 for the treatment of hypoxic term or near-term newborns with persistent pulmonary hypertension of the newborn. NO inhalation has also been widely used to treat acute pulmonary hypertension in adults and children (1–5). Other potential uses for inhaled NO, including pulmonary vasodilation in children after cardiac surgery (6) and adults with left ventricular assist device implantation (7), have not been studied in randomized, controlled trials. There are only a few reports of long-term (more than 3 months) NO inhalation for the treatment of pulmonary arterial hypertension (8, 9), probably because of the technical and logistical difficulties of such trials with the present apparatus and cylinders. NO therapy requires gas cylinders and a cylinder distribution network, a complex delivery device to regulate NO, nitrogen dioxide (NO2), and O2 concentrations, and requires trained respiratory therapy staff. For many hospitals, inhaled NO is one of the most expensive drugs used in neonatal medicine (10). The cost of providing NO therapy for 5 days to an average newborn patient with persistent pulmonary hypertension at Massachusetts General Hospital (MGH) in Boston, MA, is about $14,000. Because of the complexity of delivering NO to patients and the expense of the therapy, inhaled NO is not available in many parts of the world and is not available for outpatient use. 1 Anesthesia Center for Critical Care Research, Department of Anesthesia, Critical Care and Pain Medicine, Massachusetts General Hospital, Harvard Medical School, Boston, MA 02114, USA. 2Division of Rheumatology, Allergy and Immunology, Department of Medicine, Massachusetts General Hospital, Harvard Medical School, Boston, MA 02114, USA. *Corresponding author. E-mail: [email protected]

Several methods have been used to produce NO for biomedical purposes, including chemical methods preparing NO from N2O4 (11), as well as various electrical systems including pulsed arc (12–16), gliding arc (17), dielectric barrier (18, 19), microwave (20), corona (21), radio frequency–induced coupled discharge (22), and nonthermal atmospheric pressure high-frequency plasma discharge (23, 24). However, these methods produce large amounts of NO2 and ozone (O3) as toxic byproducts, requiring complex purification systems (12–14, 16, 23, 25, 26). In 1992, we reported a method to produce therapeutic levels of inhaled NO from electric plasma discharge in air (27). With the subsequent approval and widespread acceptance of NO as a therapeutic drug, and advances in miniaturized electronic circuitry, gas monitoring, and power storage capacity in batteries, it is readily possible to develop a simple, portable, and safe method of bedside and ambulatory NO production. Here, we sought first to develop and test a simple, lightweight, portable, and economical method of producing pure NO from ambient air or an O2-enriched mixture at the bedside and, second, to investigate whether electrically generated NO (either continuously produced or triggered during early inspiration) could reduce pulmonary hypertension in a large animal model. Our lightweight device with low power consumption differs from previous devices, which are heavy, require scavengers at high temperature, have large power requirements, and produce large amounts of NO and contaminating toxic gas (NO2). We tested the biomedical application of our NO generators in 30- to 35-kg lambs because their pulmonary circulation closely mimics the pediatric pulmonary circulation. We compared electrical plasma production of NO with NO delivered from a cylinder, which is the clinical gold standard. Electrical plasma discharge produced NO in the therapeutic range of 5 to 80 parts per million (ppm), and low levels of potentially toxic gas byproducts, such as NO2 and O3, could effectively be removed by an inline calcium hydroxide [Ca(OH)2] scavenger. In lambs with pulmonary hypertension, electrically generated NO was as effective as NO delivered from a cylinder in producing prompt pulmonary vasodilation and

www.ScienceTranslationalMedicine.org

1 July 2015

Vol 7 Issue 294 294ra107

1

Downloaded from http://stm.sciencemag.org/ on May 25, 2016

Inhalation of nitric oxide (NO) produces selective pulmonary vasodilation and is an effective therapy for treating pulmonary hypertension in adults and children. In the United States, the average cost of 5 days of inhaled NO for persistent pulmonary hypertension of the newborn is about $14,000. NO therapy involves gas cylinders and distribution, a complex delivery device, gas monitoring and calibration equipment, and a trained respiratory therapy staff. The objective of this study was to develop a lightweight, portable device to serve as a simple and economical method of producing pure NO from air for bedside or portable use. Two NO generators were designed and tested: an offline NO generator and an inline NO generator placed directly within the inspiratory line. Both generators use pulsed electrical discharges to produce therapeutic range NO (5 to 80 parts per million) at gas flow rates of 0.5 to 5 liters/min. NO was produced from air, as well as gas mixtures containing up to 90% O2 and 10% N2. Potentially toxic gases produced in the plasma, including nitrogen dioxide (NO2) and ozone (O3), were removed using a calcium hydroxide scavenger. An iridium spark electrode produced the lowest ratio of NO2/NO. In lambs with acute pulmonary hypertension, breathing electrically generated NO produced pulmonary vasodilation and reduced pulmonary arterial pressure and pulmonary vascular resistance index. In conclusion, electrical plasma NO generation produces therapeutic levels of NO from air. After scavenging to remove NO2 and O3 and filtration to remove particles, electrically produced NO can provide safe and effective treatment of pulmonary hypertension.

RESEARCH ARTICLE reducing elevated pulmonary arterial pressure (PAP) and pulmonary vascular resistance index (PVRI).

RESULTS Design and construction of an electric NO generator Here, we designed, constructed, and tested two prototypes for electrical plasma generation of NO from air or from gas mixtures containing up to 90% O2 and 10% N2. Our “offline” device produced NO by creating a spark gap between electrodes powered by a microcontroller circuit board and injected the NO gas product into the ventilator tubing

(an offline system; Fig. 1, A and B). With offline NO production, a Ca(OH)2 scavenger removed potential toxic gases (NO2 and O3) before delivery to the patient (or, in this study, the lamb). The inline device was installed in the inspiratory tubing (an “inline” system) (Fig. 1, C and D) and could produce NO either continuously or timed with the beginning of inspiration. In contrast to the offline device, which synthesized NO by spark generation offline and then pumped the NO-containing gas into the airway circuit, the inline device produced NO in the inspiratory line of the subject, which allowed a precise quantity of NO (proportional to breath volume) and synthesized NO only during inspiration (and/or just before inspiration), thus saving power, preventing degradation of the

Downloaded from http://stm.sciencemag.org/ on May 25, 2016 Fig. 1. Schematic of electric plasma NO generators. (A) Ventilatory placement for offline NO generator for animal testing. (B) Detailed internal components of the offline NO generator. Purple arrows indicate gas in and out. Black arrows indicate that the sensors and pump were connected to or controlled by the circuit board to monitor the readings via computer. Air or an O2/N2 mixture was pumped and filtered through a 0.22-mm high-efficiency particulate absorption (HEPA) air filter. The gas flow rate was measured with a meter. Sensors for O2, NO, and NO2 indicated the concentration of each gas. The electrodes were powered by a microcontroller circuit. The Ca(OH)2 scavenger and a 0.22-mm filter removed potential toxic gases (NO2 and O3) and metal particles before delivery to the lambs. (C) Position of the inline NO generator in the ventilatory pathway. (D) Inline NO generator contained iridium-platinum electrodes (2.0-mm gap, shown on left), a 0.22-mm filter, and 12 g of Ca(OH)2. (E) Illustration of electrical discharge variables for spark generation. The B and N values controlled how many hundreds or thousands of times per second the plasma discharge was repeated. The P and H values controlled individual pulse energy (duration and period of voltage and current). (F) Schematic of NO generator for bench testing. The flow of air or O2/N2 mixtures was controlled by a flowmeter, passed through an O2 sensor to indicate O2 concentration, and entered the NO generator chamber. The electrodes were powered by a microcontroller circuit board. The inline scavenger reduced the NO2 level, and NO and NO2 sensors indicated the level of each gas. All bench tests (Figs. 2 to 4) used this schematic. www.ScienceTranslationalMedicine.org

1 July 2015


2

RESEARCH ARTICLE and NO2 production by an electric spark, we measured NO and NO2 with O2 levels of 10, 21, 50, 80, and 90%, balanced with N2 entering an inline system at a constant gas flow rate of 5 liters/min. NO production varied with changing O2 and N2 levels (Fig. 2A). Maximal NO levels (68 ± 4 ppm) at 50% O2 were generated with a sparking regimen of B = 25, N = 35, P = 240, and H = 100. Lower levels of NO with this spark regimen were produced at 10, 21, 80, and 90% O2. The production of NO2 ranged from 2 to 4 ppm at the varying O2 levels. Running the NO generator continuously at 5 liters/min with 21% O2 for 10 days produced 50 ppm NO without change in the NO or NO2 levels (Fig. 2B). These results suggest that maximal NO is produced at 50% O2, and the inline NO generation system is stable in continuous operation for 10 days.

The effect of varying O2 and N2 input levels on the inline production of NO and NO2 gases An optimal NO-generating device would yield maximal NO and minimal NO2. To investigate how varying levels of O2 and N2 would affect NO

The effect of varying air pressure on the production of NO and NO2 Varying atmospheric pressure also changes NO and NO2 production (14). However, it is not clear whether altering atmospheric pressure will selectively affect NO or NO2 production (that is, the ratio of NO2/NO may vary). To study the effect of changing atmospheric pressure on NO and NO2 production, we generated electric plasma arcs at 1/3, 1/2, 1, and 2 atmospheres absolute pressure (ATA). Compared to NO and NO2 production at 1 ATA, NO and NO2 production was reduced at 1/3 and 1/2 ATA and increased at 2 ATA (Fig. 2C). The ratio of NO2/NO production was similar at each ATA tested.

Fig. 2. NO and NO2 levels produced by inline electric plasma generator. (A) Gas was produced at varying O2 and N2 levels (settings: B = 25; N = 35; P = 240; H = 100; gas flow rate, 5 liters/min). *P < 0.05 versus NO generated at 10 and 90% O2, unless otherwise indicated [one-way analysis of variance (ANOVA) with post hoc Newman-Keuls test]. (B) Gas was produced for 10 days at the following settings: B = 30; N = 20; P = 240; H = 75; airflow rate, 5 liters/min. (C) NO and NO2 were produced at varying atmospheric pressures (settings:

B = 100; N = 10; P = 140; H = 10; generation time, 1 min in a 500-ml chamber). P values were determined by one-way ANOVA with post hoc NewmanKeuls test. (D) NO and NO2 were produced with an iridium electrode covered with (+) or without (−) PTFE membrane. (E) NO2/NO ratio with different electrode compositions at an airflow rate of 5 liters/min. P values were determined by one-way ANOVA with post hoc Newman-Keuls test. All data are means ± SD (n = 3 to 5).


1 July 2015


3


electrode, and shortening the time available for oxidation of NO to NO2 in the airway. For the inline system, we designed and constructed a low gas flow resistance Ca(OH)2 scavenger for removing NO2 and O3. On the basis of the equation Ca(OH)2 + NO2 + NO → Ca(NO2)2 + H2O, and assuming that 2 ppm NO2 was produced continuously at a gas flow rate of 5 liters/min for 24 hours by electric NO generation, we estimated that it would take 45 mg of Ca(OH)2 to scavenge this quantity of NO2. A 0.22-mm air filter followed the scavenger to be certain that metal particles released from the electrodes or Ca(OH)2 particles were not inhaled. Four pulse pattern control variables were selected to precisely control the level of NO production in both inline and offline systems (Fig. 1E). These variables included the number of spark groups per second (B), the number of spark discharges per group (N), the time in microseconds between two spark discharges (P), and the pulse time in microseconds (H). B and N values controlled how many hundreds or thousands of times per second the plasma discharge was repeated. P and H values controlled individual pulse energy (voltage and current). The schematic layout of NO generation in Fig. 1F was used for all of the bench tests (results in Figs. 2 to 4).

RESEARCH ARTICLE We used a microporous polytetrafluoroethylene (PTFE) membrane to protect the inline NO spark generator from potential damage caused by droplets of moisture or airway secretions from the patient. To determine how the membrane might alter gas delivery, we measured NO and NO2 levels before and after passage through the membrane. The PTFE membrane had no significant effect on the delivery of NO or NO2 (Fig. 2D).

The effect of varying the sparking regimen on the production of NO and NO2 To optimize NO production by varying B, N, P, and H, we first changed the value of B or N while keeping the P and H values constant. As the value of B or N increased, NO and NO2 production increased linearly (Fig. 3, A and B). While maintaining B and N constant, NO and NO2 production increased when the value of P was increased up to 250 ms or when the value of H was increased up to 100 ms (Fig. 3, C and D). These data indicate that NO production can be precisely controlled and that gas production increases with increased pulse repetition rate (B and N) and energy storage capacitance (P and H). A Ca(OH)2 scavenging system removes NO2 and O3 without altering the delivery of NO NO2 is one of the potentially toxic byproducts of electrically generated NO. We used a scavenger system containing 75 g of Ca(OH)2 to remove NO2 produced by the plasma arc generator. To test the efficiency of the scavenging system when the electric NO generator was in the offline configuration, we measured NO and NO2 levels entering and exiting the scavenger. For 21, 50, and 80% O2 at an airflow rate of 5 liters/min, the Ca(OH)2 scavenger effectively removed NO2 to levels below the U.S. Environmental Protection Agency (EPA) limit of 1 ppm (28), without significantly reducing exiting NO (Fig. 4A). In contrast to synthesizing NO by plasma generation offline and then pumping the NO-containing gas into an airway circuit, we explored producing NO inline by intermittent or continuous electric generation. Producing NO in the inspiratory line allowed NO to be synthesized only during inspiration, avoided wasting NO during exhalation, and shortened the time available for oxidation of NO to NO2 in the airway. To limit NO2 levels in the inspiratory inline NO generator, we designed and tested a low gas flow resistance (0.02 cmH2O⋅min liter−1) inline scavenger containing a smaller amount (12 g) of Ca(OH)2 to remove NO2. More than 90% of the NO2 produced by the inline NO generator was removed by the inline Ca(OH)2 scavenger, which did not alter the NO level (Fig. 4B). Our data suggest that the inline scavenger, which uses sixfold less Ca(OH)2 with low gas flow resistance, efficiently removes NO2 without altering NO levels. An electrical discharge in O2 might be expected to produce O3 as a potentially harmful byproduct. Therefore, we measured O3 levels

Fig. 3. The effect of varying the sparking regimen on NO and NO2 production in air (79% N2/21% O2). (A) Varying B with others constant (N = 25, P = 240, H = 100). (B) Varying N with others constant (B = 35, P = 240, H = 100). (C) Varying P with others constant (B = 35, N = 25, H = 100). (D) Varying H with others constant (B = 35, N = 25, P = 240). For all of these studies, iridiumplatinum electrodes with 2.0-mm gap were used with an airflow rate of 5 liters/min. Data are means ± SD (n = 3 and 4).

produced by discharges at varying O2 levels and tested the ability of the Ca(OH)2 scavenger to remove O3. With an airflow rate of 5 liters/min and 21, 50, or 80% O2, less than 0.1 parts per billion (ppb) O3 was detected after passage through the Ca(OH)2 scavenger (Fig. 4, C and D). These results suggest that the plasma NO generator produces minimal amounts of O3 at all O2 concentrations tested, and that the Ca(OH)2 scavenger efficiently removes O3 to negligible levels that are well below the EPA O3 limits of 80-ppb exposure for 8 hours daily (29). Offline continuous electric synthesis of NO decreases PAP and PVRI After optimization of sparking regimen and scavenging, our inline and offline systems were tested in a large animal model of pulmonary hypertension and compared to the standard of care, NO from a tank. Lambs received a continuous infusion of the thromboxane analog U46619, which increased PAP to 28 to 30 mmHg within 30 min. NO was generated from a plasma discharge in 21, 50, or 80% O2 and then injected at 5 liters/min into the inspiratory limb of mechanically ventilated lambs. Lambs breathed 40 ppm NO produced by the electric NO generator for 4 min, then NO production and delivery were stopped, and PAP was measured for an additional 6 min. Breathing electrically generated NO from O2 rapidly reduced the PAP from ~28 to 29 mmHg before NO breathing to 24 mmHg at 1 min after breathing NO (Fig. 5A). Breathing offline electric plasma–generated NO also significantly decreased the PVRI at all O2 levels. As a positive control, lambs breathed 40 ppm NO from a cylinder in 50% O2. Our data indicate that electrically produced NO is as effective in reducing PAP and PVRI as breathing NO from a cylinder. Inline electrical plasma NO generation decreases PAP and PVRI To avoid wasting NO during exhalation and to shorten the time during which NO might oxidize to NO2 in the airway, we installed the NO


1 July 2015


4


The effect of electrode composition on NO2 production Ideally, our system would achieve a low ratio of NO2 to NO. To consider the possibility that different electrodes might decrease NO2 levels without altering NO production, we tested electrodes made or plated with iridium, nickel, carbon, and tungsten carbide at an airflow rate of 5 liters/min. The iridium electrode produced the lowest NO2/NO ratio (4.5 ± 0.5%) while maintaining constant NO production (altering B and N to achieve constant NO levels) as compared with nickel, carbon, and tungsten carbide electrodes (Fig. 2E). Therefore, an iridium electrode was selected to produce NO in subsequent in vivo studies.

RESEARCH ARTICLE

N = 25, P = 240, H = 70), respectively. (C and D) O3 produced by offline and inline electric plasma NO generators entering and exiting Ca(OH)2 scavenger. Gas flow rate, 5 liters/min; iridium-platinum electrodes with 2.0-mm gap. Data are means ± SD (n = 4 and 5 measurements per point). P values were determined by one-way ANOVA with a post hoc NewmanKeuls test.

generator inline in the inspiratory limb of the airway. We tested whether inline NO generation was as effective as offline NO generation in lowering PAP and PVRI. Lambs received a continuous infusion of U46619 to increase PAP to 28 to 30 mmHg, and then NO for inhalation was produced from air (21, 50, or 80% O2) by continuous sparking. Breathing electrically generated NO continuously for 4 min rapidly reduced the PAP from ~29 to ~21 mmHg, similar to the NO tank (Fig. 5B). PVRI was reduced from ~430 to 252 dynes·s cm−5 m−2. At all three input O2 concentrations, the PAP and PVRI decreased within a minute of commencing NO generation and increased within a minute after electric NO generation was discontinued. We found that inline, electrically generated NO was as effective at reducing PAP and PVRI as offline NO production or NO delivered from a cylinder: decrease in PAP was 7 ± 2 (inline) versus 5 ± 1 (offline) versus 5 ± 2 (cylinder) mmHg; decrease in PVRI was 155 ± 34 (inline) versus 138 ± 46 (offline) versus 146 ± 24 (cylinder) dynes·s cm−5 m−2. These changes in PAP and PVRI were not significant in comparing the different modes of NO generation (P > 0.05, two-way ANOVA with repeated measures).

within 1 min of commencing NO generation and increased within 1 min after electric NO generation was discontinued. Our data suggest that electrically generated NO within the inspiratory line and generated either continuously (Fig. 5B) or intermittently (Fig. 5C) effectively reduces pulmonary hypertension and vascular resistance in lambs.

The effect of continuous versus intermittent NO production on PAP and PVRI To save power and shorten the time for oxidation of NO to NO2, we placed the NO generator in the inspiratory limb and triggered it synchronously with the beginning of inspiratory airflow. NO was only generated and delivered during the initial 0.8 s of inspiration. In lambs with U46619-induced pulmonary hypertension, electrically generated NO from 21, 50, or 80% O2 rapidly reduced both the PAP and the PVRI (Fig. 5C). At all three O2 concentrations, the PAP and PVRI decreased

DISCUSSION We designed and tested generators that produce NO using pulsed electric arc discharges. The electric discharges produced stable levels of NO in the therapeutic range of 5 to 80 ppm at gas flow rates of 0.5 to 5 liters/min for 10 days. NO was produced from discharges in air, as well as discharges in O2-rich gas mixtures containing up to 90% O2 (10% N2). Small quantities of potentially toxic products generated by the plasma arc, including NO2, O3, and metal particles released by the electrodes, were efficiently removed by passing the gas through a Ca(OH)2 scavenger and a filter. After passage through the Ca(OH)2 scavenger, the level of NO was unchanged, whereas levels of NO2 and O3 were far below the National Institute for Occupational Safety and Health limits for breathing 8 hours/day at the workplace (28, 29). Compared with electrodes consisting of tungsten carbide, carbon, and nickel, an iridium electrode produced the lowest NO2 levels. The NO generator was tested both offline and within the inspiratory line. In lambs with pulmonary hypertension induced by a vasoconstrictor, both orientations promptly reduced the elevated pulmonary pressure and resistance. To produce pure NO in air using the electrical generator, we used a Ca(OH)2 scavenger to remove toxic byproducts from the gas stream. We


1 July 2015


5


Fig. 4. The Ca(OH)2 scavenger reduces the levels of NO2 and O3 exiting both the offline and the inline electric plasma generators. (A and B) NO and NO2 produced by offline electric plasma NO generator entering and exiting the 75-g Ca(OH)2 scavenger (settings: B = 25, N = 35, P = 240, H = 100) and NO and NO2 produced by inline electric plasma NO generator entering and exiting the 12-g Ca(OH)2 scavenger (settings: B = 35,

RESEARCH ARTICLE


1 July 2015


6


Here, we designed and tested two methods of triggering (continuous or intermittent) and positioning (offline or inline) the electric NO generator. Because NO is generated by high-voltage sparks, one strategy is to place the NO generator offline and away from the immediate patient care area. The offline NO generator can be connected to the ventilation circuit via gas tubing and can then deliver NO at a constant flow (5 liters/min). NO in air could also be injected during early inspiration in a volume proportional to airway gas flow rate. Another strategy is to place the NO generator inline, which permits a lightweight and portable device to produce NO triggered in early inspiration by synchronizing NO production with the respiratory flow. Because inline electric production of NO may be synchronized with inspiration, NO is only produced during one-quarter to one-eighth of the total respiratory cycle, requiring less power from a battery and favoring portable application. The inline NO generator simplifies NO delivery because it does not require a complex gas injection device such as the INOMAX DS (Ikaria Inc.) (30). In lambs with chemically induced pulmonary hypertension, we report that electrically generated NO, delivered either continuously or by inspirationtriggered production, produces pulmonary vasodilation and reduces the PAP and PVRI as efficiently as injecting NO from a cylinder. Fig. 5. Breathing electrically generated NO reverses pulmonary vasoconstriction in lambs. PulmoA limitation of clinical NO inhalation nary hypertension was induced by administering U46619 to lambs for 30 min before NO breathing therapy is the need to store NO in gas cy[baseline (BL)]. NO, either electrically generated or from a tank, was administered at 30 min, for 4 min. PAP and PVRI were measured every 1 min. Settings: B = 35; N = 25; P = 240; H = 100; tidal volume, 400 ml; linders, which are expensive and bulky. respiratory rate, 10 to 14 breaths/min; gas flow rate, 5 liters/min; iridium-platinum electrodes with 2.0-mm Other investigators have attempted to gap; plasma triggered either continuously or for 0.8 s upon inspiration. For the NO tank, the same level of solve this problem with smaller NO/N2 cyNO gas was produced by diluting 500 ppm NO in N2 (cylinder). (A) Offline continuous NO generation. PAP linders (31) or by using chemical sources before and after inhaling 40 ppm NO. (B) Inline continuous NO generation. PAP and PVRI before and after to derive NO (11). Recently, the results of inhaling 20 to 40 ppm NO. (C) Inline intermittent NO generation triggered by inspiration. PAP and PVRI a 16-week study, which administered NO before and after inhaling 10 to 12 ppm NO. Data are means ± SEM (n = 8 to 10 lambs per group). *P < 0.05 from a small cylinder using an injection for time 31 to 34 min versus 30 min (before NO breathing but with pulmonary hypertension), two-way device (INOpulse, Bellerophon TheraANOVA with repeated measures. peutics Inc.), were reported. Because of the large size and weight of the injection thus designed and tested two Ca(OH)2 scavengers: an offline scavenger device, compliance with this study was reported at 33% for patients weighing 75 g and a smaller portable inline scavenger weighing 12 g. not using long-term O2 therapy cylinders (31). A device deriving The inline scavenger had a shorter path length through Ca(OH)2 and NO by heating liquid N2O4 (11) was studied in 10 patients in a phase thus less airflow resistance. In addition, to prevent the inhalation of par- 2 clinical trial, but results have not been reported (32). Safety considticles into the lungs of a patient, a 0.22-mm filter was placed in series with erations including production of high levels of NO2 will be vital in the scavenger. The filter did not impede airflow through the airway and evaluating the usefulness of this approach. In our study, we chose to would therefore not interfere with a patient who was spontaneously breath- manufacture NO from air using pulsed electric discharges. We designed ing without the assistance of the ventilator. The highest level of O3 and constructed prototypes for electrical generation of NO and tested produced during electric generation of NO was 20 ppb, which is well be- them for stability of NO production for up to 10 days. In hospitals, eleclow the EPA standard for O3 exposure of 80 ppb for 8 hours/day (29). The trically generated NO could be used to replace cylinder-derived NO to Ca(OH)2 scavenger was remarkably effective in removing O3 to levels be- treat persistent pulmonary hypertension of the newborn and acute pullow 0.1 ppb. monary arterial hypertension in children and adults. If proven safe in

RESEARCH ARTICLE

MATERIALS AND METHODS Study design The aim of this study was to develop a portable electric plasma NO generator that can produce NO for inhalation therapy from air, with safe levels of NO2 and O3. We investigated the effects of O2/N2 concentration, atmospheric pressure, and electrode compositions on NO and NO2 production. We further tested and optimized NO production by changing pulse repetition and energy input, and minimized toxic gas byproducts (NO2 and O3). Animal experiments were conducted in 3- to 4-month-old lambs using protocols approved by the Institutional Animal Care and Use Committee of MGH. To evaluate the effectiveness of electrically generated NO in producing pulmonary vasodilation in lambs, a total of 34 lambs were randomly assigned to four study groups (sample size was justified by power analysis). The experimentalists were not blinded to the identity of the study groups while assaying the multiple end points of this study. The offline NO generation system and patterns of spark generation The offline electric plasma NO generator is illustrated in Fig. 1 (A and B). Air or an O2/N2 mixture was pumped and filtered through a 0.22-mm HEPA air filter (Vital Signs Inc.). The gas flow rate was measured with a meter (S-113-8, McMillan Co.). An O2 sensor (Fujikura FCX-UWC, Servoflo Corporation), NO sensor (NO-B1, Alphasense), and NO2

sensor (NO2-D4, Alphasense) indicated the concentration of each gas. The electrodes were powered by a microcontroller circuit. Energy was stored and released by an autotransformer (ignition coil, G6y, Newmotoz) and was delivered to the spark gap to create a plasma. The plasma was produced in pulses as the energy was charged and discharged from the autotransformer coil. This plasma discharge occurred in less than 1 ms. The current discharged through the gap was limited to 15 A by a resistor in series with the electrodes (ACDelco 41-101, AutoZone). The plasma temperature was controlled by the pulse sequence and duration. The Ca(OH)2 scavenger and a 0.22-mm HEPA filter removed potential toxic gases (NO2 and O3) and metal particles. The gas mixture containing NO was delivered at 5 liters/min into the inspiratory line of an anesthesia machine (Ohmeda Modulus II). The level of NO production was controlled by four pulse pattern variables (Fig. 1E): the number of spark groups per second (B), the number of spark discharges per group (N), the time in microseconds between two spark discharges (P), and the pulse time in microseconds (H). B and N values controlled how many hundreds or thousands of times per second the plasma discharge was repeated. P and H values controlled individual pulse energy (voltage and current). B, N, P, and H values for each experiment are included in individual figure captions. The NO generator either sparked continuously at 25 groups/s or was triggered for sparking for only 0.8 s at the commencement of each inspiration, as measured by the NICO Respironics airway flowmeter. The voltage output of the airway flowmeter was amplified and compared to a set threshold output of the flowmeter for controlling the timing of the spark with respiration. When the voltage crossed the set threshold, the spark sequence was triggered. Schematic of inline NO generation In contrast to synthesizing NO by spark generation offline and then pumping the NO-containing gas into an airway circuit, we explored producing NO inline (Fig. 1, C and D) using either continuous or intermittent electric discharge. The production of NO within the inspiratory line allowed NO to be synthesized only during inspiration (and/or just before inspiration), saving power, as well as shortening the time available for oxidation of NO to NO2 in the airway. To protect sparking electrodes from water droplets and mucus in the airway and to filter out particles potentially released from the electrodes, we covered the spark gap with a microporous PTFE membrane (thickness, 0.0125 mm; pore size, 12.5 mm; Goodfellow Cambridge Ltd.). We tested NO and NO2 delivery from the spark discharge generator with or without the PTFE membrane. Measurements of NO, NO2, and O3 levels We used a Sievers Nitric Oxide Chemiluminescence Analyzer 280i (GE Power and Water) to measure the level of NO. The Sievers 280i with an M&C NO2-to-NO converter (AMP Cherokee) was used to detect NO2 levels above 2 ppm. To measure NO2 at levels below 2 ppm, we used visible light (450 nm) absorption with cavity-attenuated phase shift (CAPS) technology (Aerodyne Research Inc.) (37). The CAPS NO2 monitor operates as an optical absorption spectrometer, using a blue light–emitting diode at 450 nm as a light source, a sample cell incorporating two high-reflectivity mirrors, and a vacuum photodiode detector. We used an EC 9810 series UV Ozone Analyzer (American Ecotech) to measure O3 levels. To test how varying atmospheric pressure would change NO and NO2 production, we used a sealed cylinder (500 ml) with a movable


1 July 2015


7


long-term animal inhalation studies, electric plasma–generated NO has many potential future clinical applications. Because our device is lightweight and requires only a small battery power supply and air, ambulatory electric generation of NO could be used outside of the hospital for additional indications, including the treatment of the many millions of patients with pulmonary hypertension associated with chronic lung and cardiac diseases. There have been short-term studies in patients with these diseases that showed the positive hemodynamic effects of breathing NO (33–36), but the lack of lightweight, portable NO inhalation therapy has precluded large clinical trials. The availability of a compact, lightweight, electric NO generation device will permit studies to identify uses for chronic NO inhalation therapy, including the combination of NO and systemic vasodilators for the treatment of patients with heart failure associated with pulmonary hypertension. Battery-operated NO generation would allow the use of NO therapy in places where the supply of electricity is scarce or intermittent. Another advantage of a lightweight electric device to produce NO is potential use during medical evacuation or on the battlefield, situations in which the weight of a cylinder may otherwise prevent the use of inhaled NO. In conclusion, we report that electrical plasma NO generation stably produces therapeutic levels of inhaled NO from air. After scavenging to remove NO2 and O3, electrically generated NO provides safe and effective pulmonary vasodilation in lambs with pulmonary hypertension. Electrically generated NO can be used to treat patients requiring selective pulmonary vasodilation, as is presently provided by NO in cylinders. Future studies of lightweight and portable electrical plasma generation of NO will investigate whether inhaled NO in the ambulatory patient can be used to treat chronic pulmonary hypertension and chronic obstructive pulmonary disease.

RESEARCH ARTICLE piston equipped with a pressure meter to create various atmospheric pressures (1/3, 1/2, 1, and 2 ATA) and measured NO and NO2 levels under each condition. To optimize NO production by electrical discharge while producing the lowest NO2 and O3 levels, we tested electrodes made of tungsten carbide, carbon, nickel (all from McMaster-Carr), and the noble metal iridium (ACDelco 41-101, AutoZone) using 21% O2/79% N2 at a gas flow rate of 5 liters/min.

4. 5. 6. 7.

8.

9.

10. 11.

12.

13. 14. 15. 16. 17.

18.

19.

20.

21.

Statistical analysis All values are expressed as means ± SD unless otherwise stated. All variables were normally distributed by the Shapiro-Wilk test. Data were analyzed using a one-way ANOVA with a post hoc Newman-Keuls test (GraphPad Prism, GraphPad Software Inc.). A two-way ANOVA with repeated measures was performed to determine the effect of breathing electric generated NO on PAP and PVRI with varying concentrations of O2. Probability values were two-tailed, and P < 0.05 was considered significant.

22. 23.

24.

25.

26.

REFERENCES AND NOTES 1. J. D. Roberts, D. M. Polaner, P. Lang, W. M. Zapol, Inhaled nitric oxide in persistent pulmonary hypertension of the newborn. Lancet 340, 818–819 (1992). 2. R. H. Clark, T. J. Kueser, M. W. Walker, W. M. Southgate, J. L. Huckaby, J. A. Perez, B. J. Roy, M. Keszler, J. P. Kinsella; Clinical Inhaled Nitric Oxide Research Group, Low-dose nitric oxide

27. 28. 29.


1 July 2015


8


Lamb studies Animal studies were approved by the Institutional Animal Care and Use Committee of MGH (Boston, MA). We studied 34 3- to 4-monthold Polypay lambs from a cesarean-derived, specific pathogen–free sheep flock (New England Ovis) weighing 32 ± 2 kg (means ± SD). General anesthesia was induced with 5% inhaled isoflurane (1-chloro2,2,2-trifluoroethyldifluromethyl ether, Baxter) in O2 delivered via a mask and then maintained with 1 to 4% isoflurane in 50% O2 during surgery. After tracheal intubation, animals were instrumented with indwelling carotid artery and pulmonary artery catheters. All hemodynamic measurements were performed in anesthetized lambs ventilated with a mechanical ventilator (model 7200, Puritan Bennett) at a tidal volume of 400 ml and rate of 10 to 14 breaths/min. To induce pulmonary hypertension, a potent pulmonary vasoconstrictor U46619 (Cayman Chemical), the analog of the endoperoxide prostaglandin H2, was infused intravenously at a rate of 0.8 to 0.9 mg/kg per minute to increase the mean PAP to 30 mmHg (38). The mean arterial pressure and PAP were continuously monitored using a Gould 6600 amplifier system (Gould Electronics Inc.). Pulmonary capillary wedge pressure, heart rate, and cardiac output were intermittently measured at baseline, during U46619 infusion, and before and after inhalation of NO. Cardiac output was assessed by thermal dilution as the average of three measurements after an intravenous bolus injection of 10 ml of ice-cold saline solution. PVRI and cardiac index were calculated using standard formulae. We compared the reduction in pulmonary vasoconstriction and hypertension during U46619 infusion by breathing electrically produced NO with the reduction by breathing the same level of NO gas produced by diluting 500 ppm NO in N2 delivered from a cylinder (Airgas). For studies involving continuous offline or inline NO production, we measured NO and NO2 levels by chemiluminescence. For inline inspiratory NO synthesis, we continuously sampled airway gas at 200 ml/ min and reported the peak values of sampled NO.

3.

therapy for persistent pulmonary hypertension of the newborn. N. Engl. J. Med. 342, 469–474 (2000). J. P. Kinsella, S. R. Neish, E. Shaffer, S. H. Abman, Low-dose inhalational nitric oxide in persistent pulmonary hypertension of the newborn. Lancet 340, 819–820 (1992). S. H. Abman, Inhaled nitric oxide for the treatment of pulmonary arterial hypertension. Handb. Exp. Pharmacol. 218, 257–276 (2013). N. Finer, K. J. Barrington, Nitric oxide for respiratory failure in infants born at or near term. Cochrane Database Syst. Rev., CD000399 (2006). A. M. Atz, D. L. Wessel, Inhaled nitric oxide in the neonate with cardiac disease. Semin. Perinatol. 21, 441–455 (1997). M. N. Kavarana, M. S. Pessin-Minsley, J. Urtecho, K. A. Catanese, M. Flannery, M. C. Oz, Y. Naka, Right ventricular dysfunction and organ failure in left ventricular assist device recipients: A continuing problem. Ann. Thorac. Surg. 73, 745–750 (2002). K. Vonbank, R. Ziesche, T. W. Higenbottam, L. Stiebellehner, V. Petkov, P. Schenk, P. Germann, L. H. Block, Controlled prospective randomised trial on the effects on pulmonary haemodynamics of the ambulatory long term use of nitric oxide and oxygen in patients with severe COPD. Thorax 58, 289–293 (2003). R. J. Barst, R. Channick, D. Ivy, B. Goldstein, Clinical perspectives with long-term pulsed inhaled nitric oxide for the treatment of pulmonary arterial hypertension. Pulm. Circ. 2, 139–147 (2012). N. Subhedar, C. Dewhurst, Is nitric oxide effective in preterm infants? Arch. Dis. Child. Fetal Neonatal Ed. 92, F337–F341 (2007). M. A. Lovich, D. H. Fine, R. J. Denton, M. G. Wakim, A. E. Wei, M. Y. Maslov, L. G. Gamero, G. B. Vasquez, B. J. Johnson, R. F. Roscigno, R. J. Gilbert, Generation of purified nitric oxide from liquid N2O4 for the treatment of pulmonary hypertension in hypoxemic swine. Nitric Oxide 37, 66–72 (2014). T. Namihira, S. Tsukamoto, D. Wang, S. Katsuki, R. Hackam, K. Okamoto, H. Akiyama, Production of nitric monoxide using pulsed discharges for a medical application. IEEE Trans. Plasma Sci. 28, 109–114 (2000). T. Namihira, S. Tsukamoto, D. Wang, S. Katsuki, H. Akiyama, R. Hackam, K. Okamoto, Production of nitric monoxide in dry air using pulsed discharge. IEEE 2, 1313–1316 (1999). T. Namihira, S. Katsuki, R. Hackam, H. Akiyama, K. Okamoto, Production of nitric oxide using a pulsed arc discharge. IEEE Trans. Plasma Sci. 30, 1993–1998 (2002). S. Sakai, M. Matsuda, D. Wang, T. Kiyan, T. Namihira, H. Akiyama, K. Okamoto, K. Toda, A compact nitric oxide supply for medical application. IEEE 1, 752–755 (2007). H. Hu, H. Liang, J. Li, Q. Zhao, J. He, Study on production of inhaled nitric oxide for medical applications by pulsed discharge. IEEE Trans. Plasma Sci. 35, 619–625 (2007). B. Benstaali, P. Boubert, B. G. Cheron, A. Addou, J. L. Brisset, Density and rotational temperature measurements of the OH° and NO° radicals produced by a gliding arc in humid air. Plasma Chem. Plasma Process. 22, 553–571 (2002). I. Stefanovic, N. K. Bibinov, A. A. Deryugin, I. P. Vinogradov, A. P. Napartovich, K. Wiesemann, Kinetics of ozone and nitric oxides in dielectric barrier discharges in O2/NOx and N2/O2/NOx mixtures. Plasma Sources Sci. Technol. 10, 406–416 (2001). K. Kornev, N. Yavorovsky, S. Preis, M. Khaskelberg, U. Isaev, B.-N. Chen, Generation of active oxidant species by pulsed dielectric barrier discharge in water-air mixtures. J. Int. Ozone Assoc. 28, 207–215 (2006). M. A. Wójtowicz, F. P. Miknis, R. W. Grimes, W. W. Smith, M. A. Serio, Control of nitric oxide, nitrous oxide, and ammonia emissions using microwave plasmas. J. Hazard. Mater. 74, 81–89 (2000). R. D. Hill, I. Rahmim, R. G. Rinker, Experimental study of the production of nitric oxide, nitrous oxide, and ozone in a simulated atmospheric corona. Ind. Eng. Chem. Res. 27, 1264–1269 (1988). E. Stoffels, Y. A. Gonzalvo, T. D. Whitmore, D. L. Seymour, J. A. Rees, A plasma needle generates nitric oxide. Plasma Sources Sci. Technol. 15, 501–506 (2006). S. Kühn, N. Bibinov, R. Gesche, P. Awakowicz, Non-thermal atmospheric pressure HF plasma source: Generation of nitric oxide and ozone for bio-medical applications. Plasma Sources Sci. Technol. 19, 015013 (2010). K. Matsuo, H. Yoshida, J. Choi, R. Hamid, R. Hosseini, T. Namihira, S. Katsuki, H. Akiyama, Nitric oxide generated by atmospheric pressure air microplasma. IEEE 8, 999–1003 (2009). F. Massines, P. Rabehi, P. Decomos, R. B. Gadri, P. Ségur, C. Mayoux, Experimental and theoretical study of a glow discharge at atmospheric pressure controlled by dielectric barrier. J. Appl. Phys. 83, 2950 (1998). W. J. M. Samaranayake, Y. Miyahara, T. Namihira, S. Katsuki, R. Hackam, H. Akiyama, Ozone production by pulsed power in dry air. IEEE 2, 1326–1329 (1999). W. M. Zapol, U.S. Patent 5,396,882 (1992). 1988 OSHA PEL Project Documentation (Centers for Disease Control and Prevention, Atlanta, GA, 1988). Occupational Safety Health and Administration: Ozone (U.S. Department of Labor, Washington, DC, 2012).

RESEARCH ARTICLE Acknowledgments: We thank F. Ichinose and E. S. Buys of the Department of Anesthesia, Critical Care and Pain Medicine for their critical suggestions and comments reviewing the manuscript. We also acknowledge P. Hardin and M. Hickcox of Odic Inc. for electronic design and fabrication. Funding: This study was supported by funds of the Department of Anesthesia, Critical Care and Pain Medicine, MGH. Author contributions: W.M.Z. and B.Y. designed and conceived the study and wrote and edited the manuscript; B.Y. and S.M. performed experimental work and collected, assembled, analyzed, and interpreted data; A.H.B. assisted with stability study of NO generation system; D.B.B. made major contribution revising the manuscript; and all authors gave their final approval of the manuscript. Competing interests: The MGH has filed patents on the electronic generation of NO, and several authors have a right to receive royalties. W.M.Z. is on the scientific advisory board of Third Pole Inc., which has an option to license patents on NO generators from MGH. Data and materials availability: Our electric plasma NO generator (offline or inline) is not commercially available but is available at cost via material transfer agreement (contact W.M.Z.). Submitted 15 November 2014 Accepted 3 June 2015 Published 1 July 2015 10.1126/scitranslmed.aaa3097 Citation: B. Yu, S. Muenster, A. H. Blaesi, D. B. Bloch, W. M. Zapol, Producing nitric oxide by pulsed electrical discharge in air for portable inhalation therapy. Sci. Transl. Med. 7, 294ra107 (2015).


1 July 2015


9


30. INOMAX DS Operation Manual (Ikaria Inc., Hampton, NJ, 2012), pp. 1–136. 31. R. L. Benza, R. J. Barst, A. M. Poms, J. T. Granton, B. D.B., A. Frost, H. W. Farber, N. Hirani, W. Hill, S. Shapiro, T. Fleming, R. Heyrman, C. G. Elliott, Am. Thorac. Soc. Int. Conf. 191, A2194 (2015). 32. J. Blundon, Nitric oxide, GeNO nitrosyl delivery system, NCT01165047; https://clinicaltrials. gov/ct2/show/results/NCT01165047?term=GeNO&rank=2§=X674015#more. 33. J. J. Lepore, A. Maroo, L. M. Bigatello, G. W. Dec, W. M. Zapol, K. D. Bloch, M. J. Semigran, Hemodynamic effects of sildenafil in patients with congestive heart failure and pulmonary hypertension: Combined administration with inhaled nitric oxide. Chest 127, 1647–1653 (2005). 34. O. Sitbon, F. Brenot, A. Denjean, A. Bergeron, F. Parent, R. Azarian, P. Herve, B. Raffestin, G. Simonneau, Inhaled nitric oxide as a screening vasodilator agent in primary pulmonary hypertension. A dose-response study and comparison with prostacyclin. Am. J. Respir. Crit. Care Med. 151, 384–389 (1995). 35. M. J. Semigran, B. A. Cockrill, R. Kacmarek, B. T. Thompson, W. M. Zapol, G. W. Dec, M. A. Fifer, Hemodynamic effects of inhaled nitric oxide in heart failure. J. Am. Coll. Cardiol. 24, 982–988 (1994). 36. I. Blanco, J. Ribas, A. Xaubet, F. P. Gómez, J. Roca, R. Rodriguez-Roisin, J. A. Barberà, Effects of inhaled nitric oxide at rest and during exercise in idiopathic pulmonary fibrosis. J. Appl. Physiol. 110, 638–645 (2011). 37. P. L. Kebabian, E. C. Wood, S. C. Herndon, A. Freedman, A practical alternative to chemiluminescence-based detection of nitrogen dioxide: Cavity attenuated phase shift spectroscopy. Environ. Sci. Technol. 42, 6040–6045 (2008). 38. C. Frostell, M. D. Fratacci, J. C. Wain, R. Jones, W. M. Zapol, Inhaled nitric oxide. A selective pulmonary vasodilator reversing hypoxic pulmonary vasoconstriction. Circulation 83, 2038–2047 (1991).

RESEARCH ARTICLE

Abstracts One-sentence summary: Generating nitric oxide from air by pulsed electrical discharge produces therapeutic levels of nitric oxide for inhalation and provides effective pulmonary vasodilation. Editor’s Summary: Breath of fresh NO


1 July 2015


10


Inhalation therapy—an expensive, yet life-saving, treatment for pulmonary hypertension—is not common in many parts of the world or in patients’ homes. To make inhaled nitric oxide (NO) available to all, at the bedside, Yu et al. designed a lightweight, portable system that created NO from ambient air or oxygen-nitrogen mixtures. Their generator used a pulsed electrical spark from an iridium electrode to create gaseous NO, pump it through a scavenger (to remove toxic gases like ozone), and then deliver it constantly or only during inspiration. The device was designed to have all components, including the scavenger, inline, making it self-contained and portable. The NO produced was equivalent to NO delivered from a cylinder, the current clinical gold standard, when treating lambs with pulmonary hypertension. It is hoped that this new economical technology for NO inhalation therapy will liberate patients from costly, cumbersome tanks and complicated delivery and monitoring systems, in emergency situations and in their own homes.

Producing nitric oxide by pulsed electrical discharge in air for portable inhalation therapy Binglan Yu, Stefan Muenster, Aron H. Blaesi, Donald B. Bloch and Warren M. Zapol (July 1, 2015) Science Translational Medicine 7 (294), 294ra107. [doi: 10.1126/scitranslmed.aaa3097] Editor's Summary

The following resources related to this article are available online at http://stm.sciencemag.org. This information is current as of May 25, 2016.

Article Tools

Related Content

Permissions

Visit the online version of this article to access the personalization and article tools: http://stm.sciencemag.org/content/7/294/294ra107 The editors suggest related resources on Science's sites: http://stm.sciencemag.org/content/scitransmed/5/208/208sr5.full http://stm.sciencemag.org/content/scitransmed/5/200/200ra117.full http://stm.sciencemag.org/content/scitransmed/3/88/88ra55.full http://stm.sciencemag.org/content/scitransmed/2/44/44ra58.full Obtain information about reproducing this article: http://www.sciencemag.org/about/permissions.dtl

Science Translational Medicine (print ISSN 1946-6234; online ISSN 1946-6242) is published weekly, except the last week in December, by the American Association for the Advancement of Science, 1200 New York Avenue, NW, Washington, DC 20005. Copyright 2016 by the American Association for the Advancement of Science; all rights reserved. The title Science Translational Medicine is a registered trademark of AAAS.


Breath of fresh NO Inhalation therapy−−an expensive, yet life-saving, treatment for pulmonary hypertension−−is not common in many parts of the world or in patients' homes. To make inhaled nitric oxide (NO) available to all, at the bedside, Yu et al. designed a lightweight, portable system that created NO from ambient air or oxygen-nitrogen mixtures. Their generator used a pulsed electrical spark from an iridium electrode to create gaseous NO, pump it through a scavenger (to remove toxic gases like ozone), and then deliver it constantly or only during inspiration. The device was designed to have all components, including the scavenger, inline, making it self-contained and portable. The NO produced was equivalent to NO delivered from a cylinder, the current clinical gold standard, when treating lambs with pulmonary hypertension. It is hoped that this new economical technology for NO inhalation therapy will liberate patients from costly, cumbersome tanks and complicated delivery and monitoring systems, in emergency situations and in their own homes.