Nanoparticle-mediated Photodynamic Therapy as a Method to Ablate Oral Cavity Squamous Cell Carcinoma in Preclinical Models

Abstract Photodynamic therapy (PDT) is a tissue ablation technique able to selectively target tumor cells by activating the cytotoxicity of photosensitizer dyes with light. PDT is nonsurgical and tissue sparing, two advantages for treatments in anatomically complex disease sites such as the oral cavity. We have previously developed PORPHYSOME (PS) nanoparticles assembled from chlorin photosensitizer–containing building blocks (∼94,000 photosensitizers per particle) and capable of potent PDT. In this study, we demonstrate the selective uptake and curative tumor ablation of PS-enabled PDT in three preclinical models of oral cavity squamous cell carcinoma (OCSCC): biologically relevant subcutaneous Cal-33 (cell line) and MOC22 (syngeneic) mouse models, and an anatomically relevant orthotopic VX-2 rabbit model. Tumors selectively uptake PS (10 mg/kg, i.v.) with 6-to 40-fold greater concentration versus muscle 24 hours post-injection. Single PS nanoparticle–mediated PDT (PS-PDT) treatment (100 J/cm2, 100 mW/cm2) of Cal-33 tumors yielded significant apoptosis in 65.7% of tumor cells. Survival studies following PS-PDT treatments demonstrated 90% (36/40) overall response rate across all three tumor models. Complete tumor response was achieved in 65% of Cal-33 and 91% of MOC22 tumor mouse models 14 days after PS-PDT, and partial responses obtained in 25% and 9% of Cal-33 and MOC22 tumors, respectively. In buccal VX-2 rabbit tumors, combined surface and interstitial PS-PDT (200 J total) yielded complete responses in only 60% of rabbits 6 weeks after a single treatment whereas three repeated weekly treatments with PS-PDT (200 J/week) achieved complete ablation in 100% of tumors. PS-PDT treatments were well tolerated by animals with no treatment-associated toxicities and excellent cosmetic outcomes. Significance: PS-PDT is a safe and repeatable treatment modality for OCSCC ablation. PS demonstrated tumor selective uptake and PS-PDT treatments achieved reproducible efficacy and effectiveness in multiple tumor models superior to other clinically tested photosensitizer drugs. Cosmetic and functional outcomes were excellent, and no clinically significant treatment-associated toxicities were detected. These results are enabling of window of opportunity trials for fluorescence-guided PS-PDT in patients with early-stage OCSCC scheduled for surgery.


Photophysical and photochemical characterisation
The absorbance spectra of PS and 64 Cu-PS nanoparticles were measured using UV-Vis spectrophotometry (Agilent Cary 60), and the fluorescence spectra measured using fluorescence spectrometry (Horiba FluoroMax) with excitation wavelength 416-nm.The spectra of 'intact' PS and 64 Cu-PS nanoparticles were measured in distilled water, whereas the spectra of disassembled PS monomers (i.e., Sahovaler A, Valic MS, Townson JL, et al.

Cancer Res Commun
Supplementary Methods (CRC-23-0269-AT) Page 4 of 11 porphyrin-lipid conjugate) were measured in aqueous medium containing 1.0 v/v% non-ionic surfactant (Sigma-Aldrich X100).Measurement of 64 Cu-PS samples were performed following storage at 2-8 °C until the sample radioactivity had completely decayed (e.g., ~10x half-lives of 64 Cu or ~6 days).The molar attenuation (ε) profile of intact and disassembled PS was derived from linear regression of the absorption spectra of serial diluted PS in aqueous media from 0.1-200 µM and reported on the basis of mols of porphyrin-lipid conjugate.
The structurally dependent singlet oxygen ( 1 O2) generation of PS nanoparticles was measured using singlet oxygen sensor green (SOSG) assay (Thermo Fisher S36002).SOSG reagent was freshly prepared in methanol (100 mM) and mixed with PS samples (1~2 µM porphyrin-lipid conjugate amount) before dilution in either 1x PBS (intact nanoparticle) or 1.0 v/v/% non-ionic surfactant (Sigma-Aldrich X100) to yield a final SOSG concentration of 10 µM.Replicate samples were aliquoted onto a 96-well black polystyrene microplate (Corning 3915) and illuminated with a 671-nm diode pumped solid state (DPSS) laser (LaserGlow Technologies R6710B1FX) using fluences ranging from 0.5 J/cm 2 to 10 J/cm 2 (50 mW/cm 2 rate constant).Negative control wells containing SOSG only in 1x PBS or 1.0 v/v/% non-ionic surfactant were included.Following laser illumination, the fluorescence of unquenched SOSG was quantified with a plate reader (BMG Labtech CLARIOstar Plus) using 488-nm excitation wavelength and 525-nm emission wavelength.Note there is no PS fluorescence within this emission window (see spectra in Figure 2D).Fluorescence readings were corrected for negative controls and the statistics from five independent experiments reported.

Whole blood and plasma pharmacokinetics
In tumour-bearing mice, the blood clearance of PS (10 mg/kg, 400-500 MBq 64 Cu/kg, IV) was assessed in cohorts of 5 mice/model using serial sampling from the saphenous vein at 9~10 timepoints postinjection.At each timepoint, approximately ~0.025 mL of whole blood was drawn from the vein with a heparinised capillary tube (Fisherbrand 41B22362566).The 64 Cu radioactivity in 0.010 mL aliquots of whole blood and of centrifuge-separated plasma were assayed using a gamma counter (PerkinElmer Wallac WIZARD 3" 1480).The activity of 64 Cu was decay-corrected to the time of injection and concentration of PS reported as percent injected dose per mL (%I.D./mL) or as µg/mL after multiplying by the injected dose of porphyrin-lipid conjugate. 64Cu is a surrogate measure for PS concentration in vivo owing to the stable 64 Cu-PS chelation under physiological conditions (see Supplementary Figure 1G).
The plasma pharmacokinetics of PS (10 mg/kg, IV) was assessed in healthy and VX-2 tumourbearing rabbits (6x rabbits total) using serial sampling from the marginal ear vein at 8~10 timepoints postinfusion.At each timepoint, approximately ~0.500 mL of whole blood was drawn from the vein with a heparinised capillary tube (Sarstedt 20.1345.100)and the plasma separated with centrifugation.Replicate

and 64 Cu-PS nanoparticles are detailed in Supplementary Table 1. Measurement of 64 Cu-PS samples were performed following storage at
Sahovaler A, Valic MS, Townson JL, et al.PEGylated porphyrin-lipid conjugate containing nanoparticles] (PS) nanoparticles (NanoMedicine Fabrication Center) were supplied as a sterile ready-to-use product suspended in 1x phosphate buffered saline with pH 7.2.64Cu-PORPHYSOMES ( 64 Cu-PS) nanoparticles were prepared by radiolabelling the pre-formed PS nanoparticles with positron-emitting Copper-64 ( 64 Cu) metal in a "onepot" reaction: 64 Cu(II)Cl2 (Washington University School of Medicine) supplied in 0.1 M HCl solution was diluted with 0.2 M sodium acetate (Sigma-Aldrich 32319), pH 5.5 solution before being added to PS nanoparticles with volume ratio 1:1.The specific activity of 64 Cu-PS nanoparticle at the time of labelling was ~ 49.7 ± 13.9 ×10 9 Bq/mol (on basis of mols of porphyrin-lipid conjugate).The mixture was heated to ~55 °C for 60 minutes and then chilled on ice after labelling.The labelling efficiency of 64 Cu-PS was Postnova CF2000): the retention times of replicate nanoparticle samples were analysed using a constant field method varying the field strength between 1373 g (2,250 rpm) and 2,692 g (4,500 rpm).The average effective molar mass (M50) was derived assuming a spherical particle model with a ~5-nm thick bilayer (from TEM measurements) and an aqueous-filled core.pHmeasurements were performed on PS and 64 Cu-PS samples without dilution and using a micro pH electrode (Thermo Scientific 9810BN).The hydrodynamic diameter and polydispersity index (PDI) of PS and 64 Cu-PS dilute in distilled water were measured in a glass cuvette (Malvern Panalytical PCS1115) using dynamic light scattering (DLS) with a 532-nm wavelength 'green' laser(Malvern Panalytical Zetasizer   Nano).The zeta (ζ)-potential of PS and 64 Cu-PS dilute in distilled water were measured in a dip cell change in purity normalised to the initial sample purity at 0 h.Regression analysis of data from 1 h to 48 h was performed using a one phase decay model and plateau = 0 % (i.e., complete 64 Cu-PS dissociation at time infinity) to estimate the dissociation kinetics of 64 Cu from 64 Cu-PS nanoparticles in PBS or 50 v/v% FBS.
confirmed using instant thin layer chromatography (iTLC): replicate aliquots of 64 Cu-PS were dilute 1:4 in 0.05 M ethylenediaminetetraacetic acid (EDTA) (Sigma-Aldrich E9884) and then separated on heat activated silica gel plates (Agilent SGI0001) using 0.05 M EDTA as the chromatography solvent.The PSbound 64 Cu remains at the origin (Rf 64 Cu-PS ~0.14) whereas the unbound/free 64 Cu metal migrate with the solvent front (Rf free 64 Cu > 0.88).After development, the iTLC plate was cut in two and the top (free 64 Cu) and bottom ( 64 Cu-PS) sections measured with a gamma counter (PerkinElmer Wallac WIZARD 3" 1480). 64Cu-PS nanoparticles were visualised with transmission electron microscopy (TEM) (FEI Tecnai) on formvar coated grids (Electron Microscopy Sciences FCF400-Cu-UB) after negative staining with 2.0 w/v% uranyl acetate (Electron Microscopy Sciences 224002).The lipid composition of PS and 64 Cu-PS were determined with ultra-performance liquid chromatography-evaporative light scatting (UPLC-ELS, Waters ACQUITY): replicate aliquots of PS and 64 Cu-PS were diluted in methanol (Sigma-Aldrich 154903) and passed through a BEH C18 reverse-phase chromatography column (Waters 186002350) running a mobile phase consisting of an acidic ammonium acetate (Supelco AX1222) and methanol mixture.The molecular weights of PS and 64 Cu-PS nanoparticles were measured in 1x phosphate buffered saline (PBS) (Corning 21-031-CV) carrier buffer by centrifugal field-flow fractionation-multi-angle light Sahovaler A, Valic MS, Townson JL, et al.Cancer Res Commun Supplementary Methods (CRC-23-0269-AT) Page 3 of 11 scatting/dynamic light scattering (cFFF-MALS/DL) (PS and 64 Cu-PS nanoparticle physiological stability experiments The stabilities of PS and 64 Cu-PS nanoparticles were evaluated under simulated physiological conditions: particle samples were dilute in 50 v/v% foetal bovine serum (FBS) (Gibco 12483020) (or in 1x PBS as a control) and maintained at 37 °C with constant shaking.For size stability analysis, PS samples at each incubation timepoint were filtered through a dextran-agarose Superdex 200 gel column (GE Health Care Life Sciences 17-1043-02) to remove unbound serum proteins (i.e., high molecular weight PS nanoparticles have shorter elution times than low molecular weight serum proteins).PS samples were collected from the column and the sample PDI measured using DLS.For 64 Cu chelation stability analysis, 64 Cu-PS samples at each incubation timepoint were dilute 1:4 in 0.05 M EDTA and the free 64 Cu separated from PS-bound 64 Cu using iTLC assay for labelling efficiency described previously.An experimental control for possible serum protein-bound 64 Cu was included at each timepoint.The radiochemical purity of 64 Cu-PS samples (corrected as needed for any protein-bound 64 Cu) was measured and reported at each timepoint as percent