Supplementary Materials for

stm.sciencemag.org/cgi/content/full/13/580/eabe3889/DC1

Supplementary Materials for

Percutaneous liquid ablation agent for tumor treatment and drug delivery

Hassan Albadawi, Zefu Zhang, Izzet Altun, Jingjie Hu, Leila Jamal, Kelly N. Ibsen, Eden E. L. Tanner, Samir Mitragotri*, Rahmi Oklu*

*Corresponding author. Email: [email protected] (R.O.); [email protected] (S.M.) Published 10 February 2021, Sci. Transl. Med. 13, eabe3889 (2021)

DOI: 10.1126/scitranslmed.abe3889

The PDF file includes:

Materials and Methods

Fig. S1. Characterization of concentration-dependent LATTE diffusion in the rat liver.

Fig. S2. Ex vivo microCT scans of the N1S1 liver tumor after treatment with saline,

EtOH, or LATTE.

Fig. S3. Effect of intratumoral injection of LATTE on recruitment of macrophages

and T lymphocytes.

Fig. S4. Time-dependent evaluation of drug diffusion and retention after LATTE injection into

the rat liver.

Fig. S5. The cytotoxic effect of combined LATTE and Dox on HepG2 cells.

Fig. S6. Effect of LATTE treatment on PANC-1 and SNU-478 cancer cell viability.

Fig. S7. Image-guided injection of LATTE into the pig liver.

Fig. S8. NIRF imaging and histological evaluation of explanted human tumors after injection

with LATTE.

Fig. S9. NIRF imaging and histological evaluation of explanted human renal cancer tumors after

injection with LATTE.

Table S1. Serum blood chemistry in rats with N2S1 tumors receiving intratumor injection of

saline, EtOH, or LATTE.

Legend for data file S1

Legend for movie S1

References (51–56)

Other Supplementary Material for this manuscript includes the following:

(available at stm.sciencemag.org/cgi/content/full/13/580/eabe3889/DC1)

Data file S1 (Microsoft Excel format). Individual subject-level data.

Materials and Methods

Intraparenchymal injection of LATTE into normal rat liver

Intrahepatic injections of LATTE mixtures were performed by laparotomy in 18 anesthetized Sprague Dawley rats (Envigo, Placentia, CA) placed in a supine position on a warming platform. A cotton-tipped applicator dipped in sterile saline and blunt tweezers were used to position the left lower liver lobe to receive two injections of LATTE mixtures 2 cm apart delivered with syringes with 25-gauge needles. The medial injection site received 50 µL injection of 25% LATTE mixed with 0.25 mg/mL ICG in normal saline; whereas the lateral injection site received 50 µL injection of 25% LATTE, 0.25 mg/mL ICG, and 50 µg Doxorubicin (Dox, Pfizer Pharmaceutical Company, NY) in normal saline. Following injection, the subcutaneous tissue and dermis layer were closed using 5-0 vicryl sutures. Subgroups of 6 rats were housed for 1, 7, or 28 days following injections. At the end of experimental period, rats were euthanized and explanted livers were fixed, and ex vivo fluorescent imaging was performed to measure the fluorescent intensity of ICG and Dox and to calculate the area of diffusion at each injection site.

Histopathology and immunohistochemistry

Harvested tissues were fixed in 10% buffered formalin and then axially transected to expose the core of the treatment zone. Fluorescent scans of each tissue were acquired before and after transection, and tissue was subsequently embedded in paraffin. Paraffin-embedded blocks were sectioned at 4 μm and mounted on positively charged glass slides (Fisher Scientific, #12-550-15, Pittsburgh, PA). Slides were baked at 60°C for 30 min, deparaffinized (2 × 100 % Xylene, 3 × 100 % ethanol, 1 × 95 % ethanol, 1 × 80 % ethanol, 1 × 70 % ethanol, 5 min each) and rehydrated in double-distilled water for 5 min. Serial sections were stained with hematoxylin and eosin (H&E) or Picrosirius red staining (Mastertech stain kits, CA) to visualize tissue morphology and cellular infiltration. Other sections underwent immunohistochemistry staining as previously described (51). Antigen retrieval was performed in 10 mM sodium citrate, pH 6.0 solution (Life Technologies #005000, Thermo-Fisher Scientific, Waltham, MA). Hydrated tissue sections were permeabilized in 0.1 % Triton X-100 (Sigma T8532, St. Louis, MO) in PBS and endogenous peroxidase activity was quenched with 0.3% H2O2 in 60% methanol for 30 min at room temperature. Endogenous biotin activity was blocked using

Avidin-Biotin blocking reagents (Life Technologies, #004303, Thermo-Fisher Scientific, Waltham, MA). Non-specific antibody binding was reduced by incubating in 5% goat serum in PBS for 1 h at room temperature. To identify actively proliferating cells and cells

undergoing apoptosis, tissue sections were incubated with 1:250 dilution of rabbit IgG specific for proliferating cell nuclear antigen (PCNA, AB13847, Abcam, MA) or an IgG that recognizes cleaved caspase-3 (1:250, AB13847, Abcam, MA), respectively. Rat anti-mouse CD3 IgG3к (1:20, 550295, BD Pharmingen, CA) was used to visualize naïve T lymphocytes; and polyclonal rabbit anti-CD68 IgG (1:250, AB125212, Abcam, MA) was used to recognize local monocytes and macrophages. Sections were then incubated with horseradish peroxidase-conjugated IgG (1:300, AB97051, Abcam, MA) secondary antibodies at room temperature for 30 min. Permanent insoluble dark brown color (3, 3’-diaminobenzidine, DAB substrate, Vector Laboratories SK4100, Burlingame, CA) was developed under a light microscope. Tissue sections were then counterstained with Hematoxylin (Sigma, GHS-216, St. Louis, MO) for 10 s. Slides were dehydrated 1 x in 70 % ethanol, 1 × 80 % ethanol, 1 × in 95 % ethanol, 1 × 100 % ethanol, and 100% xylene for 30 s, respectively. Next, the slides were dried and mounted with a permanent mounting solution (Histomount Solution, #008030, Life Technologies). An EVOS FL Auto-2 microscope was used to obtain digitally stitched micrographs at 200× magnification. Stitched images were loaded into ImageJ software (ver. 1.52s, National Institute of Health, USA) and deconvoluted using the IHC-profiler plugin as previously described (52, 53). Measurements were performed in the region of interest (ROI) using manual tracing in each section. Appropriate thresholds and particle size values were applied to all tumor sections to calculate the cross-sectional area and the number of positive cells in the ROI. Data were expressed as the mean of positive cells per square millimeter.

Serum collection and analysis in rats

Blood samples were collected through a needle puncture of the inferior vena cava using a 21-gauge needle syringe before euthanasia. The blood samples were allowed to clot for 30 minutes at room temperature and then centrifuged at 1,500g for 5 minutes in a vacuette tube (Greiner Bio-One, North American, Inc., Monroe, NC). Serum aliquots were stored at -80°C until analysis. Serum amounts of alkaline phosphatase (ALP) and alanine aminotransferase (ALT), creatinine (CRE), blood urea nitrogen (BUN), glucose (Glu) were measured using DRI-CHEM 4000 analyzer (Heska, Loveland, CO). C- reactive protein (CRP) was measured using quantitative ELISA (ThermoFisher Scientific, Waltham, MA), according to the manufacturer’s instructions.

Cytotoxicity assay

HepG2 human liver cancer cell line (CRL10741, American Type Culture Collection, Manassas, VA), the ampulla of Vater, adenocarcinoma (SNU-478) cells, and pancreatic carcinoma of ductal origin (PANC-1) cells (CRL-1469, American Type Culture Collection, Manassas, VA) were cultured in 75-cm2 flasks using growth medium consisting of Dulbecco's Modified Eagle Medium (DMEM, ThermoFisher Scientific, Waltham, MA) and 10% heat-inactivated bovine serum that was supplemented with 100 IU penicillin and 10 μg/mL streptomycin (Thermo Fisher Scientific, Waltham, MA). Cells were incubated inside a 5% CO2 incubator at

37°C until they reached confluency; the cells were detached using 0.05% Trypsin-EDTA solution (Millipore-Sigma, Burlington, MA) and seeded into 96-multiwell replicate plates at 5,000 cells per well and grown for 24 hours. Following the 24-hour seeding period, the medium was replaced with 200 µL of fresh growth medium containing serially diluted LATTE or Dox into designated replicate wells, and the cells were incubated for 24 or 48 hours. At the end of the incubation period, the medium was removed, and the wells were rinsed three times with Dulbecco’s modified phosphate buffer (DPBS, Sigma-Aldrich, Saint Louis, MO) followed by the addition of 100 µL growth medium. Cytotoxicity of LATTE and Dox on HepG2 was determined by adding 10 µL of freshly prepared solution of the water soluble 2-(4-Iodophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium sodium reagent (WST-1, Cayman

Chemicals, Ann Arbor, MI) into each well followed by 2-hour incubation inside a 5% CO2 humidified incubator at 37°C. WST-1

output was assessed by measuring the optical density at a wavelength of 450 nm using a microplate reader (SpectraMax iD5,

Molecular Devices, San Jose, CA). Cell viability was calculated relative to the untreated control wells that received aliquots of growth medium alone. Viability rate was calculated as follows: Viability (%) = (1-ODtreatment/ODcontrol) × 100%. The fractional viability

dose-response plots for LATTE or Dox were used to calculate the concentrations that induce cytotoxicity in 50% of the HepG2 cells (IC50)

at 24 or 48 hours based on individual treatment using the Prism Software ver. 8 (GraphPad, San Diego, CA).

In vitro evaluation of cytotoxicity to combined therapy

To establish a cytotoxic synergy effect of combined LATTE and Dox on human cancer cells, HepG2 cells were seeded into 96-multiwell replicate plates at 5,000 cells per well and grown for 24 hours. Cells were then incubated with a single agent using 4 – 5 concentrations spanning the IC50 of LATTE or Dox. Subsequently, pairwise treatments of all possible combinations were generated to

yield a 4 × 5 matrix for each singly tested drug; the values of the pairwise concentrations were interpolated from the fitted Hill curves of each single treatment dose-response curve using the Loewe model and Loewe synergy plots using Combenefit software (54, 55). All reported values are the means of three replicate experiments with each study having 8 wells per dose.

Percutaneous injection of LATTE mixture under US-guidance into pig liver

US-guided injection of LATTE solution was performed in euthanized pigs. Two mL of 25% LATTE and 0.25 mg/mL ICG mixture loaded into a syringe (Becton-Dickenson, Franklin Lakes, NJ) on a 7-cm long and 21-gauge access needle (COOK Medical,

Bloomington, IN). Ultrasonography was used to visualize the left liver lobe parenchyma using a high-frequency transducer (A 9-MHz multi-frequency linear probe, ACUSON S2000, Siemens, Germany). The access site position was marked on the skin and a scalpel blade was used to create a small incision in the center of the marked line. The access needle was advanced through the skin incision in front of the transducer until it reached the desired location, then the LATTE mixture was slowly injected over one minute. At 10 minutes post-injection, the liver tissue was harvested and subjected to NIRF imaging (IVIS 200, PerkinElmer, Inc. Waltham, MA), followed by fixation and histologic evaluation.

Magnetic resonance imaging of pig liver

Magnetic resonance (MR) imaging was performed on explanted pig liver following subcapsular injection of 2 mL solution comprised of 25% LATTE and 0.25 mg/mL ICG using a 21-gauge vascular access needle. Pig livers were scanned with a 3T MAGNETOM Skyra MRI (Siemens Healthcare, Erlangen, Germany) with an 18-channel anterior coil in combination with a 32-channel posterior spine coil. Several MR scans were performed. For anatomical reference the following scan was done: Coronal T2 single-shot fast spin-echo (HASTE) with FOV 300 × 300mm, resolution 448 × 310, slice thickness 1.3 mm, TR 800 ms, TE 120 ms, BW 620 Hz/Px, 4 NEX, scan time 3:15 min. To visualize the liver parenchyma, the following scan was done: High-resolution coronal 3D T1 fast gradient echo volumetric scan (MPRAGE) with FOV 320 × 320mm, resolution 512 × 512, slice thickness 0.6 mm, TR 1350 ms, TE 2.34 ms, Flip angle 9°, TI 900 ms, BW 390 Hz/Px, scan time 6 min resulting in 0.6 × 0.6 × 0.6 mm 3D volumetric resolution. To visualize vessels, the following scan was done: High-resolution coronal 3D T2 fast spin-echo volumetric scan (SPACE) with FOV 300 × 300mm, resolution 320 × 320, slice thickness 0.9 mm, TR 1700 ms, TE 105 ms, Flip angle 135°, BW 600 Hz/Px, NEX 1.4, scan time 9:24 min resulting in 0.9 × 0.9 × 0.9 mm 3D volumetric resolution. A coronal T2 cine fast steady-state free precession (TRUFI) scan was performed with FOV 380 × 380mm, resolution 256 × 256, slice thickness 5 mm, TR 630 ms, TE 11.77 ms, Flip angle 60°, BW 1300 Hz/Px, scan time 3:09 min resulting in a temporal resolution of approximately 1.5 frames/sec. Scans were acquired at 0 and 90 minutes after injection. Segmentation and volume calculations were acquired using the Materialise 3-Matic and Mimics 3D image processing software (Materialise, Belgium).

Fig. S1. Characterization of concentration-dependent LATTE diffusion in the rat liver. (A) Photograph of neat ionic liquid

(choline and geranic acid) in a vial and LATTE containing ICG and Exitron in a syringe. Below the image is the chemical structure of the ionic liquid used in the study. (B) Representative image showing subcapsular injection of 25% LATTE mixed with ICG and Exitron into rat liver. (C) Image of rat liver at 24 hours following injection of 50%, 25%, or 6.25% LATTE. Scale bar = 1 cm. (D) 3D rendering of reconstructed microCT images of 50%, 25%, or 6.25% LATTE injection sites in rat liver visualized with Exitron. Scale bar ꞊ 2 mm. (E) 3D volumetric analysis of the segmented microCT data showing volume of Exitron diffusion resulting from 25%-LATTE compared to 50% or 6.25%-LATTE (n ꞊ 5). (F) Viscosity measurements of 6.25%, 25%, and 50% LATTE mixtures at a shear rate of 50 1/s (n ꞊ 3). (G) Representative injection force plots generated using 6.25%, 25%, 50%, and 100% LATTE loaded into a 1 mL syringe and extruded through a 21-gauge percutaneous access needle. Data are representative of 3 experiments. (H) Images of explanted and fixed liver tissues at 24 hours following injection of 50%, 25% or 6.25% LATTE showing treatment zone (white arrows). Scale bar = 1 cm. (I to K) H&E-stained rat liver sections obtained at 24 hours following intraparenchymal injection of 50%, 25%, or 6.25% LATTE mixed with ICG and Exitron showing areas of coagulative necrosis that transitions to interstitial edema and cellular swelling along the periphery (dashed outline). High magnification images show disrupted tissue architecture, nuclear loss, scattered pooling of erythrocytes, and granulocyte infiltration. *: P < 0.05; **: P < 0.01; ****: P < 0.0001 were calculated using one-way ANOVA and Tukey’s multiple comparison tests. Data are the mean ± SEM.

Fig. S2. Ex vivo microCT scans of the N1S1 liver tumor after treatment with saline, EtOH, or LATTE. Images of 3D-rendered

microCT scans following segmentation showing N1S1 tumor in rat livers at two weeks following injection with saline, ETOH, or LATTE. Red indicates hyper densities inside the tumor; pink color outlines the tumor volume. Scale bar = 0.5 cm.

Fig. S3. Effect of intratumoral injection of LATTE on recruitment of macrophages and T lymphocytes. (A to C) Representative

histology images of immunostained sections visualizing CD68+ cells obtained from saline-, ETOH-, or LATTE-treated N1S1 tumors. (D) Quantitative analysis of CD68+ cell count in N1S1 histology sections in the tumors injected with the indicated reagent (n = 6). (E to G) Representative immunostained histology sections visualizing CD3+ T lymphocytes in saline-, ETOH-, or LATTE-treated N1S1 tumor sections obtained at two weeks following injection. (H) Quantitative analysis of CD3+ cell count in sections of tumor injected with the indicated reagent (n = 6). **: P < 0.01; ****: P < 0.0001; statistical significance determined by ANOVA with Tukey’s post hoc comparisons for D or Dunnett’s post hoc test for H. Data in graphs represent the mean ± SEM.

Fig. S4. Time-dependent evaluation of drug diffusion and retention after LATTE injection into the rat liver. (A) A panel of

fluorescent images of explanted rat livers that received a subcapsular injection of 50 µL 25% LATTE mixed with ICG or 25% LATTE mixed with ICG and doxorubicin at two locations in the same liver lobe. Images were acquired 1, 7, and 28 days after injection. (B) Summary of measured ICG diffusion area using near-infrared scanning of the liver surface (n = 12). (C) Quantitative analysis of average fluorescence radiance of doxorubicin at the injection sites at 1, 7, and 28 days following injection (n = 6). (D) A panel of serial histology sections of rat liver stained with H&E or picrosirius. Tissues were obtained at 1, 7, and 28 days following LATTE injection. Black arrow indicates tissue remodeling associated with fibrosis as depicted by the intense red stain of collagen. Data are representative of 1 of 6 livers in each injection condition. Scale bars = 2000 µm. ns: P > 0.05; *: P < 0.05; **: P < 0.01 was calculated using ANOVA with Tukey’s post hoc tests. Data presented as the mean ± SEM.

Fig. S5. The cytotoxic effect of combined LATTE and Dox on HepG2 cells. (A) Dose-response curve of cells exposed to serial

dilutions of LATTE for 24 or 48 hours. (B) Dose-response curve of cells exposed to serial dilutions of doxorubicin for 24 or 48 hours. (C and D) Synergy plot derived from a dose-response matrix showing a synergistic cytotoxic effect on HepG2 cells at 24 and 48 hours following treatment with LATTE and doxorubicin (n = 12). Data represent replicates of three independent experiments (n = 12).

V

ia

b

ili

ty

(

%

)

Dox

_(µM

)

LA

TT

E %

D

C

24 Hours

V

ia

b

ili

ty

(

%

)

Dox

_(µM

)

LA

TT

E %

48 Hours

Fig. S6. Effect of LATTE treatment on PANC-1 and SNU-478 cancer cell viability. Plot showing dose-dependent decrease in

fractional viability of PANC-1 (IC50 of 0.3%) and SNU-473 cells (IC50 of 0.21%) at 24 hours following LATTE treatment. Data

represent replicates of three independent experiments (n = 12).

-2 -1 0 1 2 0.00 0.25 0.50 0.75 1.00 1.25

Log (LATTE %)

F

ra

c

ti

o

n

a

l

V

ia

b

il

it

y

Fig. S7. Image-guided injection of LATTE into the pig liver. (A) Photograph demonstrating percutaneous image-guided injection

of 25% LATTE (white arrow). (B) Corresponding US image during LATTE injection showing the injection needle (white arrow) and the echogenic appearance of LATTE (yellow dashed outline) during injection as it spreads within the liver parenchyma. (C) NIRF image of explanted pig liver lobe shows a robust fluorescent signal corresponding to the location of the LATTE injection site. Scale bar = 1 cm. (D) NIRF and histologic image of liver sections obtained from the LATTE injection site localizing the fluorescent area (left) to the ablation zone (right, black dashed outline). (E) High magnification micrograph of H&E-stained histology section obtained from the LATTE-treated zone (* denote the location in panel D, right). (F) Representative T1-weighted magnetic resonance (MR) sequence of pig liver obtained 90 minutes following LATTE injection (green outline). (G) MR volumetric analysis showing a 2.8-fold increase in the circumferential diffusion of LATTE within 90 minutes following LATTE injection. (H) NIRF scan of the LATTE-injected liver in (F) showing ICG diffusion in the treated zone. * P < 0.05 was calculated using a paired t-test. Data are the mean ± SEM (n = 4).

Fig. S8. NIRF imaging and histological evaluation of explanted human tumors after injection with LATTE. A panel of resected

human tumors including: (A) colorectal cancer liver metastasis (CRCLM), (B) hepatocellular carcinoma (HCC), (C)

cholangiocarcinoma (CCA), (D) breast cancer (BC), (E) Metastatic appendiceal cancer (MAC), (F) lung adenocarcinoma (LAC), and (G) synovial sarcoma (SS). Each panel (left to right) contains a gross view image of each tissue, NIRF images obtained 10 minutes and 24 hours following LATTE injection, H&E-stained histology sections from LATTE-treated (left) and untreated (right) tissue. Scale bar for histology images = 200 µm and for gross images = 1 cm.

Fig. S9. NIRF imaging and histological evaluation of explanted human renal cancer tumors after injection with LATTE. (A to E) A panel of five resected human renal cell carcinoma tumors (RCC). Each panel (left to right) contains a gross view image, NIRF

images obtained 10 minutes and 24 hours following LATTE injection, H&E-stained histology sections from LATTE-treated (left) and untreated (right) tissue. Scale bar for histology images = 200 µm and for gross images = 1 cm.

Table. S1. Serum blood chemistry in rats with N2S1 tumors receiving intratumor injection of saline, EtOH, or LATTE.

Summary of serum amounts of alkaline phosphatase (ALP), alanine aminotransferase (ALT), creatinine (CRE), C-Reactive protein (CRP), blood urea nitrogen (BUN), glucose (Glu), and total protein at two weeks following intratumoral injection with saline, ETOH, or LATTE. There was no statistical difference in all measured values between the groups. All measured values in the experimental rats remained within the normal limits of healthy Sprague Dawley rats (56). Statistical analysis was calculated using one-way ANOVA. Data are presented as the mean ± SEM (n = 6).

Analyte Saline ETOH LATTE P Value

ALP U/L 199 ± 17 188 ± 28 211 ± 13 0.69 ALT U/L 62 ± 12 48 ± 4.7 57 ± 6 0.48 CRE mg/dL 0.35 ± 0.02 0.38 ± 0.08 0.31 ± 0.02 0.65 CRP µg/mL 472 ± 39 425 ± 58 459 ± 38 0.27 BUN mg/dL 26 ± 2 26 ± 3 25 ± 1 0.99 GLU mg/dL 220 ± 21 203 ± 58 268 ± 34 0.49 Total protein g/dL 5.4 ± 0.2 5.6 ± 0.1 5.2 ± 0.2 0.4

Data file S1. Individual subject-level data. (Separate file)

Movie S1. Image-guided injection of LATTE into the rabbit VX2 liver tumor. Video demonstrates percutaneous needle-based