Effect of Air Pockets in Drug Delivery in Jet Injections

Home , Depot injection, Jet injector

bioRxiv preprint doi: https://doi.org/10.1101/2021.02.02.429451; this version posted February 3, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Eﬀect of air pockets in drug delivery in jet injections

Pankaj Rohilla, Emil Khusnatdinov, and Jeremy Marston∗

Department of Chemical Engineering, Texas Tech University, Lubbock, TX 79409

Needle-free jet injections are actuated by a pressure impulse that can be delivered by different mechanisms, and the resultant jets are O(102) m/s. Here, we report on the effect of entrapped air bubbles since filling procedures for pre-filled ampoules can induce bubbles, especially for viscous fluids. We use spring-piston devices as the principal actuation mechanism and vary both the location and size of the initial bubble. We find that the bubble location does have a statistically significant (p < 0.05) effect on the jet exit speed, based upon the volumetric flow rate. However, we reveal subtle features such as intermittent atomization when the gas pockets pass through the orifice and de-pressurize, which leads to spray formation and a temporary increase in jet dispersion, both of which can lead to product loss during an injection. These results have implications for the development of prefilled ampoules for jet injection applications.

INTRODUCTION the advantages of prefilled ampoules is the low variability in the amount of drug without any air Conventional oral administration of drugs is pocket [6]. However, air pockets can form within limited by low bioavailibility caused by the the liquid inside the ampoules during loading or acidic environment in the stomach and restric- transportation, which can further be amplified tive intestinal epithelium [1]. The most com- by the viscosity effects of the contained fluid. mon alternative to oral drug delivery are hy- Such air pockets can affect the volume and other podermic needle syringes, which have numer- variables related to drug delivery. The effect of ous caveats including needle-stick injuries, cross- air pockets on the efficacy of drug delivery is still contamination, and needle-phobia [2,3,4] and largely unknown. This demands a systematic potential limitation on injectability of viscous parametric study to further the current under- drugs. Thus, evidently there is a need for an standing. effective needle-free technique for drug admin- In general, the presence of air pockets within a istration. Needle-free jet injections are one of drug is either undesirable or could be harnessed the alternative techniques where a high-speed in several medical applications[7,8,9, 10, 11]. narrow jet (diameter, dj ∼100-200 µm) punc- Ultrasound imaging [12, 13], lithotripsy [14], tures the top layer of the skin (i.e. stratum controlled cavitation in gene delivery [13] and corneum) and deposits the drug into the intra- laser-induced jet injection [15, 16, 17] are some dermal/subcutaneous/intramuscular tissue [5]. of the applications where energy generated due Hypodermic syringes and needle-free jet injec- to the formation and collapse of air pockets can tors are medical devices which need an ampoule be utilized. On the other hand, the phenomenon or barrel to store medication or biological drugs of collapsing air pocket also generates microjets before injection. These ampoules are either sup- and shock waves, which can rupture red blood plied prefilled or can be loaded manually. One of cells in artificial heart valves and can be detri-

1 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.02.429451; this version posted February 3, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

mental to soft tissue [18, 19, 20]. tail in other works [33, 34, 35], but comprises a Introducing air bubbles (0.1-0.2 ml) before stiff spring piston with a cartridge with an orifice drawing medication into a syringe was common- of diameter, do∼157 µm. DI water was filled in place practice among clinicians and nurses in a transparent nozzle cartridge with a volume ca- the United States before the widespread use of pacity of 0.11 ml. Air pockets were introduced at disposable syringes in the 1960s [6, 21]. These different locations inside the nozzle cartridge (as air bubbles were used to correct the medica- shown in figure 1(a)) using 1 ml luer-lock syringe tion dosage to account for the dead volume in with 24-gauge needle. Furthermore, the plunger the needle hub. Moreover, air pockets when in- was carefully replaced to avoid the escape of air jected with the medication were thought to pre- pockets via the orifice exit. We have used five vent the seepage or backflow of the medication different locations within the nozzle cartridge to into the subcutaneous layer through the needle understand their effect on the bubble dynamics tracks [22]. Multiple studies showed contradic- with time. These locations were selected on the tory results to this hypothesis [23, 24]. Moreover, basis of their probability of occurrence. For lo- air bubbles occupy the available volume for the cation L1, an air pocket was introduced at or in medication which cause dosage inaccuracy [25]. close proximity of the plunger tip. It should also Thus, it is recommended to avoid air bubbles in be noted that the bubble locations, L2 and L3 syringe injections [6, 21, 25]. are isotropic radially. In jet injection, air bubbles can affect the co- To capture the bubble and jet dynamics, a herency and continuity of the jet stream in addi- high-speed video camera (Phantom V711, Vi- tion to dosage errors. This phenomenon intensi- sion Research Ltd.) was used at frame rates fies with an increase in plunger speed. Thus, in in a range of 10,000-30,000 frames per second. the context of jet injectors where the jet speeds The tip of the plunger was tracked to get the can be O(102) m/s, it is important to study plunger displacement (figure 1(b)) with time the effect of air pockets in the nozzle on drug using Photron FASTCAM Analysis (PFA ver. delivery via jet injection. Although researchers 1.4.3.0 ) software. Plunger speed was then esti- have studied the effect of cavitation in the stor- mated from the slope of the displacement-time age and injection of therapeutics in the past plot after the initial ringing phase, e.g. t > 5 [7, 8, 26, 27, 28, 29, 30, 31, 32], there is a lack ms (figure 1(c)). Furthermore, we estimated the of detailed study to understand the effect of air jet speed at the orifice exit, vj (figure 1(d)) from pockets in intradermal injections via jet injec- plunger speed using mass conservation, assuming tors. the liquid to be incompressible at the experimen- Here, we study the effect of air pockets present tal conditions. at different locations within the nozzle cartridge A single side-view perspective was sufficient to on the hydrodynamics and efficacy of drug deliv- track the plunger tip. However, to understand ery via jet injection. The physical properties of the bubble dynamics and for qualitative analysis the liquid and nozzle dimensions were kept con- of bubbles within the nozzle during a jet injec- stant. We studied the atomization of the bubbles tion, we employed two cameras (second camera: within the liquid and the dispersion of the jet Phantom Miro 311, Vision Research Ltd.) with when bubbles exit through the orifice. In addi- orthogonal views. In addition, using extension tion, ex vivo studies were conducted on porcine tubes with Nikon micro-nikkor 60 mm lenses, we skin to understand the effect of air pockets on achieved typical effective pixel sizes in a range of the delivery efficiency. 10-30 µm/px. Volume fraction was used to quantify the size MATERIAL AND METHODS of the air pockets, which can be defined as a ratio of the volume occupied by an air pocket (Vb) and A spring-based jet injector (Bioject ID pen) was the total liquid volume inside a nozzle cartridge used in the experiments, which is described in de- without any air pocket (V ). We measured the

2 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.02.429451; this version posted February 3, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Figure 1: Air pockets in a nozzle cartridge. (a) Five different locations of air pockets inside a nozzle cartridge, (b) trajectory (in red) of plunger motion inside the nozzle for different locations of air pockets, (c) plunger-displacement with time for a jet injection with and without air pockets, and (d) volume fraction (Vb/V ) and jet speed (vj) for different locations of air pockets. Scale bar represents 3 mm (a,b).

volume fraction occupied by air pockets at differ- jet injection. We cut skin across the center of the ent locations by weighing the liquid-filled nozzle injection site to visualize the dispersion of the with and without air pockets. Figure 1(d) shows liquid after jet injection. Trypan Blue (Sigma the volume fraction of air pockets introduced at Aldrich) was used as a dye (1 mg/ml) in DI wa- different locations inside the nozzle. ter to aid in the visualization of skin blebs, and a custom Matlab script was used to estimate the To measure the force profile during the jet in- dimensions of skin blebs. One must note that jection, a miniature load button cell (Futek-LLB although fresh porcine skin is a close model to 130, 50 lb, FSH03380 ) was placed at a distance human skin at high humidity conditions [36], we of 2 mm away from the orifice exit of the nozzle used skin after a single freeze-thaw cycle. to avoid contact with the nozzle during jet injec- We performed one-way ANOVA tests to check tion. Force profiles of jet injection were recorded the statistical significance of various parameters at a sample rate of 4,800 Hz. used in the study with a significance level of α To conduct ex vivo studies, porcine skin was = 0.05. used as a skin model for human skin. Porcine skin patches (thickness ∼3-5 mm) were har- RESULTS AND DISCUSSIONS vested from the side regions of Yorkshire-Cross pigs (age: 13 weeks), euthanized in the depart- Bubble Dynamics ment of Animal Sciences (Texas Tech Univer- sity). These skin patches were kept in a -80◦C Air pocket collapses with the inertial impact of freezer, but thawed to room temperature before the plunger, atomizing into multiple microbub-

3 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.02.429451; this version posted February 3, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Figure 2: Snapshots of bubble dynamics for diﬀerent locations of air pockets inside the nozzle cartridge. Snapshots of bubble dynamics corresponding to diﬀerent time frames for location L1(a − i) ≡ [0, 0.1, 0.2, 0.27, 0.37, 1.5, 2.27, 5.73, 61.07] ms; for location L2(a − i) ≡ [0, 0.2, 0.3, 0.5, 1.0, 2.9, 4.57, 6.23, 61.77 ms; for location L3(a − i)v ≡ [0, 0.1, 0.2, 0.27, 0.37, 1.5, 2.27, 5.73, 61.07] ms; for location L4(a − i) ≡ [0, 0.67, 0.167, 0.43, 0.67, 1.47, 2.23, 3.5, 64.67] ms; for location L5(a − i) ≡ [0, 0.5, 0.6, 0.67, 0.73, 1.1, 1.83, 5.3, 30.33] ms.

bles. We captured the qualitative behavior of ringing phase caused by the sudden impact of these micro bubbles with time as the plunger the spring piston on the plunger, followed by pushes the liquid out via a narrow oriﬁce. As a nearly linear and stable motion. Average jet shown in ﬁgure 1(c), plunger motion exhibits a speeds were calculated from the slope of this lin-

4 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.02.429451; this version posted February 3, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

ear region of the plunger displacement-time plot. and L4), microbubbles exit the nozzle at an early Introduction of air pockets did not alter the jet stage whereas for air pockets introduced farther speed (vj) significantly which lied in a range of upstream, not all of these bubbles exit; it can ∼142-146 m/s except for the case of air pocket be noticed from ith frames in figure 2 that mi- introduced at the center of the nozzle cartridge crobubbles were present in the tapered section of (L5). the nozzle at the end of injection for air pocket Among the different locations of air pockets, introduced at locations L1, L2, and L5. L5 showed a distinctive behavior. The plunger moved ∼3.7 mm within a short period of 1.5 ms, thus resulting in a rapid collapse of the air pocket. The collapsing air pocket generates a pressure opposing the plunger motion, and the interplay of forces during this period is mani- fested by the recoiling trajectory of the plunger, as shown in the last frame of figure 1(b). The impulse is much stronger, and resistance in the Figure 3: Orthogonal views of nozzle car- early stages is much weaker, leading to signif- tridge with microbubbles inside the liquid cor- icant over-pressure (noted by a force increase responding to time, t = 5.5 ms for air pocket of ∼33%). This resulted in a prolonged re- introduced at location L2. Scale bar represents 3 coil/ringing phase. Due to this extended ring- mm. Note: A vertical line on the left frame is an ar- tifact of the nozzle and is not related to the contained ing phase, we calculated the average jet speed fluid or air pockets. by fitting a line after this phase to the plunger displacement-time plot. For L5, average jet Figure3 shows orthogonal views of microbub- speed was lower (128.5±3.8 m/s). Another im- bles inside the nozzle at same instant (t = 5.5 portant characteristic of air pockets introduced ms). Different views shown in figure3 yield dif- at different locations inside the nozzle was the ferent bubble size distributions at the same in- volume fraction of the air pocket, which was stant of time due to the lensing effects of the nearly half at location L5. In the case of other nozzle geometry. Thus, to avoid the erroneous locations, the air pockets occupied ∼ 20% of the measurement, we did not quantify the bubble total available volume. size distribution with time. Figure 2 shows a montage of frames summa- rizing the bubble dynamics for air pockets placed Assuming the air bubble to be an ideal gas, at different locations. When the plunger is trig- we can estimate the volume of the wedge-shaped gered, the fluid pressure in the cartridge in- bubble cloud formed after the collapse of an air creases rapidly up to O(10 MPa), causing the pocket at location L5, using ideal gas law ( PV = air pocket to collapse and disintegrate into mi- nRT = const.): crobubbles. The trajectory of such microbubbles largely depends on the pressure gradient, which further depends on the flow disturbance caused P1V1 = P2V2 (1) by the bubbles collapsing under high pressure. Although numerous studies have been conducted Where P1 and P2 are atmospheric pres- to understand the behavior of bubbles with in- sure (∼105 Pa) and pressure inside the bubble ertial impact [29, 30, 37], the underlying physics (∼5×107 Pa, using peak force of F ∗ ∼1 N), re- behind the bubble dynamics may vary on the spectively. V1 and V2 are initial volume of air basis of applications. pocket (∼55 µL) and the volume of the wedge- The life cycle of microbubbles depends on the shape bubble cloud after the collapse, respec- location where the air pocket was introduced ini- tively. Minimum height of the bubble cloud tially. For air pockets present near the outlet (L3 (hmin) was estimated using:

5 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.02.429451; this version posted February 3, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Figure 4: Impact force proﬁles for jet injections for diﬀerent locations of air pockets. L0 represents the jet injection without any air pocket inside the nozzle. Shaded region represents ringing phase during jet injection. n=3.

∼16.9 ms, respectively. In the absence of an air pocket, the time duration of a jet was ∼38 ms. It V2 hmin ≈ 2 (2) is noteworthy that the time duration of injection Di (π 4 ) was lower in displacement-time plot (figure 1(c)) due to the difficulties associated with tracking Where Di is the inner diameter of the nozzle barrel. the plunger tip towards the end of injection. A slight asymmetry in the plunger causes a wedge-shaped air volume that disintegrates from left to right in the image sequence L5(a-i), there- As the spring piston strikes the plunger, a ring- fore, we can expect a size distribution with ing phase can be observed in force profiles in the rmin 10 µm and rmax 10 µm, evidenced initial stage (∼10 ms) of liquid injection. In the by zoomed images where bubbles with size of absence of any air pockets, the peak force was O(100 µm) can be easily seen. Moreover, many ∼0.63 N and a similar peak force was observed bubbles with a size of the order of 1 pixel or less for air pockets present at locations L1, L2, L3, were also observed. and L4 with values of ∼0.62 N, ∼0.76 N, ∼0.67 N, and ∼0.78 N, respectively. However, L5 ex- Force measurements hibited a higher force of ∼1.03 N, with an increase of 63%. After the peak force, the magni- To measure the impact force, liquid jets were im- tude of the force profile was nearly the same for 2 pinged on a load cell placed at a distance of 2 mm all cases with F¯ ' 0.4 N (F¯ = ρvj Ao, where Ao away from the orifice exit. Figure 4 shows force is the cross-sectional area of orifice exit) except profiles of jets with air pockets present at differ- the case of L5 with lower magnitude of the mea- ent locations. The time duration of the jets for sured force. Further implications of force pro- air pocket locations L1, L2, L3, L4 and L5 were file and injection duration will be discussed in ex ∼37.5 ms, ∼34.8 ms, ∼35.8 ms, ∼33.5 ms, and vivo studies.

6 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.02.429451; this version posted February 3, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Figure 5: Eﬀect of bubbles on the jet coherency with orthogonal views. (a) liquid jet without any bubbles (t = 1.4 ms), (b) bubble cloud in tapered section before exit (t = 9.9 ms), (c) spray-like jet formation as the bubbles exit with the jet (t = 10 ms), and (d) diminishing atomization of jet as bubble clouds exits the tapered section (t = 10.1 ms). Scale bar: 2 mm

Eﬀect of microbubbles on jet

As pressurized microbubbles exit the orifice, they undergo a rapid depressurization and expand, resulting in a spray-like jet (figure 5). The magnitude of the effect of microbubbles on the jet depends on the size and the number of bubbles exiting the nozzle orifice along the liquid jet. Figure 5 shows the effect of a bubble cloud exiting the nozzle on the liquid jet with orthogonal views captured using two high-speed cameras at a frame rate of 10,000 fps. The liquid jet free of any effect of bubbles is shown in fig- Figure 6: Cross section view of skin blebs for: ure 5(a), where orthogonal views show that the (a) no air pockets, (b) air pocket at location L1, (c) jet looks coherent from one side and slightly dis- air pocket at location L2, (d) air pocket at location L , (e) air pocket at location L and (f) air pocket persed from the other side. In the next frame 3 4 at location L . Scale bar represents 5 mm. (figure 5(b)), a cloud of air pockets (highlighted 5 in dashed yellow outlines) appeared in the tapered section of the nozzle with spray-like jets porcine skin harvested from different pigs (hence, as the bubbles exit with the jet. As more bubble different colors in figure 6) was within 3 mm. cloud exits the nozzle, the jet dispersion grows The effect of force exerted by the jet injector as shown in figure 5(c). With the majority of air (loading) in the normal and axial direction dur- pockets exiting the orifice exit, the remnant ef- ing injection has been reported in an earlier fects of the bubbles on the liquid jet can be seen study [35], and therefore we used a recommended in figure 5(d), but the jet will resume the steady normal load of 1 kg for the maximum delivery ef- stream upon clearance of air bubbles. ficiency at which the injection was actuated. After impingement, the liquid jet creates a Ex vivo studies hole in the skin through which liquid propagates inside the skin. The stiff poro-elastic structure of Dyed water was injected into the porcine skin the dermal tissue resists the liquid inflow inside to understand the effect of air pockets on the the skin, thus, limits the amount of liquid that percentage delivery. Porcine skins were thawed can be delivered with a single injection. To visu- to room temperature before administering in- alize the dispersion of liquid inside skin after jet jections. The thickness of the dermis layer of injection, the skins were cut across the point of

7 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.02.429451; this version posted February 3, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

injection. Figure6 shows cross-sections of skin and L5 showed higher percentage delivery of liq- blebs corresponding to the different location of uid inside the skin. Effect of varying the location air pockets. of air pockets on the delivery efficiency of dyed water was significant (p < 0.05). It is noteworthy that the delivery efficiency was affected by the confluence of changing location and available liquid volume inside the nozzle. As discussed earlier, the volume of liquid inside the nozzle changes with introduction of air pockets inside the nozzle. Introducing an air pocket at location L1 showed higher percentage delivery than the case of no air pockets for nearly the same force profile. This increase in the percentage delivery could be done due to the reduced liquid volume to be injected inside the skin. As shown in earlier studies, higher percentage delivery was obtained for jet injection of liquid in lower volumes [35]. It was hypothesized that there is a limit of liquid volume which can be injected into the skin without significant rejec- tion. Here, with the introduction of air pockets, the available liquid volume to be injected was lower for locations L1and L5, resulting in higher efficiency. In addition, the percentage delivery was nearly the same for location L3 and L4 as compared to control case of L0. Bubbles exit early with a jet for air pockets present at locations L3 and L4. As the high-speed liquid jet penetrates skin, it creates a channel to facilitate the further delivery of incoming liquid. Any disturbance in the jet Figure 7: Effect of air pockets on effective- shape in the initial phase could have an adverse ness of drug delivery. (a) delivery efficiency and projected area of bleb cross-sections for jet injections effect on the channel formation for liquid propa- with air pockets at different locations (n=5 ) and (b) gation, thus resulting in lower percentage deliv- aspect ratio of the blebs (AR = d/w), where d and w ery and narrow dispersion patterns, as observed are depth and width of skin bleb respectively (n=10 ). for L3 and L4. Higher peak force also helps in increasing percentage delivery. In the case of air Figure7 shows the effect of the location of pocket introduced at location L5, higher percent- the air pocket on the delivery efficiency and age delivery was observed; partly due to high im- the dimensions of the bleb formed inside the pact force (∼1.03 N) and lower available volume skin after jet injection. Delivery efficiency (η = of liquid to be injected (∼0.044 ml). (m−mr)×100/m) was measured from the weight The effect of location of air pockets was signif- of liquid rejected at the top of the skin after jet icant on the projected area (p < 0.05) and the injection (mr) and total available volume in noz- aspect ratio of skin blebs (p < 0.05). Projected zle cartridge (m). Whatman filter papers were area showed nearly the same trend as of deliv- used to absorb the rejected liquid on the top of ery efficiency. In case of injection with an air skin. pocket present near the exit (L3 and L4), skin Presence of air pockets at locations L1, L2, blebs showed large intrasample variation in their

8 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.02.429451; this version posted February 3, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

shape and their dimensions. potential of transdermal drug delivery. Na- ture reviews Drug discovery, 3(2):115–124, 2004. CONCLUSIONS [2] Bruce G Weniger and Mark J Papania. Al- We investigated the effect of introducing air ternative vaccine delivery methods. Vac- pockets at various locations within the liquid cines, page 1200, 2013. contained in a nozzle cartridge for jet injectors. Air pocket collapses and forms microbub- [3] Elisabetta Rapiti, A Pruss-Ustrun, and bles which affects the shape of the liquid jet. Air Y Hutin. Sharps injuries: Global burden pockets at different locations affected the peak of disease from sharps injuries to health- force during the initial ringing phase. Further- care workers. World Health Organization, more, ex vivo studies conducted on porcine skin Geneva, 2005. showed nearly the same or higher percentage delivery of dyed water inside skin. We conclude [4] Robert M Jacobson, Avril Swan, Adedunni that the high percentage delivery obtained for Adegbenro, Sarah L Ludington, Peter C air pockets at different locations was due to lower Wollan, Gregory A Poland, Vaccine Re- available volume to be injected and higher peak search Group, et al. Making vaccines more force. Moreover, the projected area and aspect acceptable—methods to prevent and mini- ratio showed a significant effect of various loca- mize pain and other common adverse events tions of air pockets in addition to a significant associated with vaccines. Vaccine, 19(17- effect on percentage delivery. Although the in- 19):2418–2427, 2001. troduction of air pockets helped in enhancing the [5] Joy Schramm-Baxter and Samir Mitragotri. delivery efficiency of the drug, air pockets should Needle-free jet injections: dependence of jet be avoided in the nozzle as it causes dosage inac- penetration and dispersion in the skin on curacy which can alter the efficacy of the vaccine. jet power. Journal of Controlled Release, The measurement of the force profiles showed 97(3):527–535, 2004. higher peak forces in the initial ringing phase due to hammer effect. Furthermore, ex vivo stud- [6] Suzanne C Beyea and Leslie H Nicoll. Ad- ies conducted on porcine skin showed similar or ministration of medications via the intra- higher percentage delivery of dyed water inside muscular route: an integrative review of the skin. We conclude that the high percent- the literature and research-based protocol age delivery obtained for air pockets could be at- for the procedure. Applied nursing research, tributed to the lower available volume and higher 8(1):23–33, 1995. peak force. In addition, the project area and aspect ratio showed a significant effect for vari- [7] Christopher E Brennen. Cavitation and ous locations of air pockets in addition to a sig- bubble dynamics. Cambridge University nificant effect on percentage delivery. Although Press, 2014. the introduction of air pockets helped in enhancing the delivery efficiency of the drug, air pock- [8] Christopher Earls Brennen. Cavitation in ets should be avoided in the nozzle as it causes medicine. Interface focus, 5(5):20150022, dosage inaccuracy which can alter the effect of 2015. the vaccine. [9] Hua Tang, Chiao Chun Joanne Wang, Daniel Blankschtein, and Robert Langer. REFERENCES An investigation of the role of cavitation in low-frequency ultrasound-mediated trans- [1] Mark R Prausnitz, Samir Mitragotri, and dermal drug transport. Pharmaceutical re- Robert Langer. Current status and future search, 19(8):1160–1169, 2002.

9 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.02.429451; this version posted February 3, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

[10] Ahmet Tezel, Ashley Sens, and Samir Mi- [18] LA Garrison, TC Lamson, Steven Deutsch, tragotri. Investigations of the role of cav- DB Geselowitz, RP Gaumond, and JM Tar- itation in low-frequency sonophoresis using bell. An in-vitro investigation of prosthetic acoustic spectroscopy. Journal of pharma- heart valve cavitation in blood. The Journal ceutical sciences, 91(2):444–453, 2002. of heart valve disease, 3:S8–22, 1994.

[11] Ahmet Tezel and Samir Mitragotri. Inter- [19] Peter Johansen. Mechanical heart valve cav- actions of inertial cavitation bubbles with itation. Expert review of medical devices, stratum corneum lipid bilayers during low- 1(1):95–104, 2004. frequency sonophoresis. Biophysical journal, 85(6):3502–3512, 2003. [20] Edmond Rambod, Masoud Beizaie, Michael Shusser, Simcha Milo, and Morteza Gharib. [12] Lawrence Crum, Michael Bailey, Joo Ha A physical model describing the mecha- Hwang, Vera Khokhlova, and Oleg Sapozh- nism for formation of gas microbubbles nikov. Therapeutic ultrasound: Recent in patients with mitral mechanical heart trends and future perspectives. Physics Pro- valves. Annals of biomedical engineering, cedia, 3(1):25–34, 2010. 27(6):774–792, 1999.

[13] Martin JK Blomley, Jennifer C Cooke, [21] Leslie H Nicoll and Amy Hesby. Intramus- Evan C Unger, Mark J Monaghan, and cular injection: an integrative research re- David O Cosgrove. Microbubble contrast view and guideline for evidence-based prac- agents: a new era in ultrasound. Bmj, tice. Applied nursing research, 15(3):149– 322(7296):1222–1225, 2001. 162, 2002.

[14] Dahlia L. Sokolov, Michael R. Bailey, and [22] S Quartermaine and R Taylor. A compar- Lawrence A. Crum. Use of a dual-pulse ative study of depot injection techniques. lithotripter to generate a localized and in- Nursing Times, 91(30):36, 1995. tensified cavitation field. The Journal of the Acoustical Society of America, 110(3):1685– [23] M Ipp, M Goldbach, S Greenberg, and 1695, 2001. R Gold. Effect of needle change and air bubble in syringe on minor adverse re- [15] Carla Berrospe Rodr´ıguez, Claas Willem actions associated with diphtheria-tetanus Visser, Stefan Schlautmann, David Fernan- toxoids-pertussis-polio vaccination in in- dez Rivas, and Ruben Ramos-Garcia. To- fants. The Pediatric infectious disease jour- ward jet injection by continuous-wave laser nal, 9(4):291–293, 1990. cavitation. Journal of biomedical optics, 22(10):105003, 2017. [24] Liam Mac Gabhann. A comparison of two depot injection techniques. Nursing Stan- [16] Akihito Kiyama, Nanami Endo, Sennosuke dard (through 2013), 12(37):39, 1998. Kawamoto, Chihiro Katsuta, Kumiko Oida, Akane Tanaka, and Yoshiyuki Tagawa. Vi- [25] Gail Chaplin, Harriet Shull, and Pual C sualization of penetration of a high-speed Welk III. How safe is the: Air-bubble tech- focused microjet into gel and animal skin. nique. Nursing, 15(9):59–62, 1985. Journal of Visualization, 22(3):449–457, 2019. [26] Donn Sederstrom. Cavitation in pharmaceutical manufacturing and shipping. 2013. [17] Pankaj Rohilla and Jeremy Marston. Feasi- bility of laser induced jets in needle free jet [27] Theresa Trummler, Spencer H. Bryngelson, injections. International Journal of Phar- Kevin Schmidmayer, Steffen J. Schmidt, maceutics, 589:119714, 2020. Tim Colonius, and Nikolaus A. Adams.

10 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.02.429451; this version posted February 3, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Near-surface dynamics of a gas bubble col- jet injections in diﬀerent skin models. Jour- lapsing above a crevice. Journal of Fluid nal of Drug Delivery Science and Technol- Mechanics, 899:A16, 2020. ogy, 60:102043, 2020.

[28] Jean-Christophe Veilleux, Kazuki Maeda, [36] SA Ranamukhaarachchi, S Lehnert, Tim Colonius, and Joseph E Shepherd. SL Ranamukhaarachchi, L Sprenger, Transient cavitation in pre-filled syringes T Schneider, I Mansoor, K Rai, UO Häfeli, during autoinjector actuation. 2018. and B Stoeber. A micromechanical comparison of human and porcine skin before and [29] Maya Mounir Daou, Elena Igualada, after preservation by freezing for medical Hugo Dutilleul, Jean-Marie Citerne, Javier device development. Scientific reports, Rodr´ıguez-Rodr´ıguez, Stéphane Zaleski, 6(1):1–9, 2016. and Daniel Fuster. Investigation of the collapse of bubbles after the impact of a piston [37] Javier Rodr´ıguez-Rodr´ıguez, Almudena on a liquid free surface. AIChE Journal, Casado-Chacón, and Daniel Fuster. 63(6):2483–2495, 2017. Physics of beer tapping. Physical review letters, 113(21):214501, 2014. [30] S Li, R Han, AM Zhang, and QX Wang. Analysis of pressure field generated by a collapsing bubble. Ocean Engineering, 117:22– ACKNOWLEDGMENTS 38, 2016. This work was financially supported by The [31] Ryota Oguri and Keita Ando. Cavitation National Science Foundation via award CBET- bubble nucleation induced by shock-bubble 1749382. interaction in a gelatin gel. Physics of Flu- ids, 30(5):051904, 2018. AUTHOR CONTRIBUTIONS

[32] Gaoming Xiang and Bing Wang. Numeri- P.R. and J.M. designed the experiments. E.K. cal investigation on the interaction of pla- and P.R. conducted the experiments. P. R. an- nar shock wave with an initial ellipsoidal alyzed the data and wrote the manuscript. J.M. bubble in liquid medium. AIP Advances, reviewed and edited the manuscript. 8(7):075128, 2018.

[33] Pankaj Rohilla, Yatish S Rane, Idera Lawal, COMPETING INTERESTS Andrew Le Blanc, Justin Davis, James B All the authors declare no competing interests. Thomas, Cormak Weeks, Whitney Tran, Paul Fisher, Kate E Broderick, et al. Char- acterization of jets for impulsively-started needle-free jet injectors: Inﬂuence of ﬂuid properties. Journal of Drug Delivery Sci- ence and Technology, 53:101167, 2019.

[34] Pankaj Rohilla and Jeremy O Marston. In- vitro studies of jet injections. International journal of pharmaceutics, 568:118503, 2019.

[35] Pankaj Rohilla, Idera Lawal, Andrew Le Blanc, Veronica O’Brien, Cormak Weeks, Whitney Tran, Yatish Rane, Emil Khusnatdinov, and Jeremy Marston. Load- ing eﬀects on the performance of needle-free