A platform technology for ultra-long acting intratumoral therapy

Intratumoral (IT) therapy is a powerful method of controlling tumor growth, but a major unsolved problem is the rapidity that injected drugs exit tumors, limiting on-target exposure and efficacy. We have developed a generic long acting IT delivery system in which a drug is covalently tethered to hydrogel microspheres (MS) by a cleavable linker; upon injection the conjugate forms a depot that slowly releases the drug and “bathes” the tumor for long periods. We established technology to measure tissue pharmacokinetics and studied MSs attached to SN-38, a topoisomerase 1 inhibitor. When MS ~ SN-38 was injected locally, tissues showed high levels of SN-38 with a long half-life of ~ 1 week. IT MS ~ SN-38 was ~ tenfold more efficacious as an anti-tumor agent than systemic SN-38. We also propose and provide an example that long-acting IT therapy might enable safe use of two drugs with overlapping toxicities. Here, long-acting IT MS ~ SN-38 is delivered with concurrent systemic PARP inhibitor. The tumor is exposed to both drugs whereas other tissues are exposed only to the systemic drug; synergistic anti-tumor activity supported the validity of this approach. We propose use of this approach to increase efficacy and reduce toxicities of combinations of immune checkpoint inhibitors such as αCTLA-4 and αPD-1.


Fluorescein labeled microspheres (FLs-MS-Ac)
Stable linked fluorescein labeled tetra-PEG hydrogel MSs with acylated amines (FLs-MS-Ac) containing the -CN α-Lys crosslink cleavage rate modulator were produced from amino-MS and suspended in isotonic acetate tween buffer (IAT, 10 mM pH 5 acetate, 145 mM NaCl, % tween 20) as previously described (1).The fluorescein concentration of the suspension (0.8 mM) was determined by absorbance by dissolving 0.050 mL of suspension into 0.450 mL of 50 mM NaOH then measuring absorbance at A495.These MS were used in experiments when FLs-MS were mixed with other MS-drug conjugates to serve as a tracer for MS at the injection site.RP was synthesized from Rhodamine B as previously reported (2).

Dual labeled microspheres (FLs-MS~RPR 2A, 2B and 2C)
The following procedure (Scheme S3) was used to synthesize azido-linker-RP MS conjugates 2A, 2B, and 2C.A slurry of FLS-MS-CO (0.024 mmol CO, 1 equiv) in acetonitrile (8 mL) was treated with one of the azido-linker-RP containing reaction mixtures (A, B, or C, 1 mL, 0.0075 mmol, azido-linker-RP, 0.3 equiv), methanol (5 mL), and acetic acid (0.001 mL).The resulting slurries were kept at room temperature for 36 h then centrifuged at 4500 x G to separate the MSs from supernatant.The MSs were then washed with 1 x 40 mL of methanol, separated by centrifugation at 4500 x G then treated with a solution of O-(2-Azidoethyl)heptaethylene glycol (0.5 mL, 40 mM, 0.020 mmol, 0.8 equiv) in water, to cap unreacted cyclooctynes.After 18 h the MSs were washed with 3 x 40 mL methanol, then with 4 x 40 mL of IAT buffer centrifuging at 4500 x G in between washes.Finally the FLs-MS~RPR slurries were transferred to 10 mL syringes and packed at 3000 x G for 5 min using a special fixture for centrifugation of the syringes (6) to provide ~5 mL of FLs-MS~RPR slurry for each material (2A, 2B, and 2C).Measurements of the RP release rates as well as the FLS and RP content for these materials are described below (Fig. S2, Fig. S3, and Table S1).

3.6
In vitro RP release rates from FLs-MS~RPR Release of RP was measured using a previously reported assay (6).In brief: samples (0.150 mL) of FLs-MS~RPR (2A, 2B, and 2C) slurry were placed into nylon mesh pouches and then the pouches were heat sealed.The pouches were placed into divided cuvettes containing 2.7 mL of pH 8.4 bicine buffer (0.1 M) at 37 o C and a magnetic stir bar.The absorbance of the buffer was measured over 35 h using a heated and stirred auto sampler on a UV-VIS spectrophotometer (Hewlett Packard 8453).RP (565 nm) was released over the course of monitoring, and as expected, no release of the stably linked FLS (495 nm) was observed.The time to reverse gelation (tRG) for the fluorescein labeled gels is 200 h at pH 8.4 (4) and should not occur over the duration of this experiment.Half-lives for RP release were determined using a first order exponential fit of the experimental data (Fig. S2).Half-lives measured at pH 8.4 were converted to pH 7.4 values based on T1/2_7.4= T1/2_8.4*10 (8.4-7.4)(3).The glass syringe was fitted with a 30 g x 1/2" needle (BD 305106) then purged by expelling all but 0.025 mL of material through the needle.Four syringes containing 0.025 mL of MSs were prepared for each material (Fig. S3A).The needle was capped with a small piece of platinum cure silicone cord (McMaster Carr 9808K21) to prevent leakage and evaporation prior to use.To assess the dosing accuracy the material in the syringes (0.025 mL) was injected into a 1.5 mL centrifuge tube containing 1.0 mL of sodium hydroxide (0.1 N).After 1 hour the absorbance of the solution was measured to detect FLs (495 nm) and RP (565 nm).The molar ratio of FLS to RP was determined for each material using A495 and A565 (Table S1).Dosing syringes for use in intra-tumoral injections were prepared in the exact same manner in a laminar flow cabinet, and using syringes that had been disinfecting with 70% ethanol.Separation of RP from FLS-MS-Ac: Two separation methods (centrifugation and filtration) were tested to isolate FLs-MS-Ac and free RP.Both methods began by injecting 0.10 mL of slurry into IAT buffer (1.0 mL), the resulting suspension was homogenized with a tissue blender (IKA T25 Ultra-Turrax homogenizer) for 1 minute at 8000 RPM Centrifugation Method: Homogenized suspensions were centrifuged 20000 x G for 2 min, then the supernatant extract was removed by pipette.The pellet of FLs-MS-Ac was washed with 5 x 1 mL of IAT, separating MSs by centrifugation at 20000 x G for 2 min between washes.
Filtration method: Homogenized suspensions were filtered using 0.2 µm PTFE spin filters (Millipore Ultrafree MC hydrophilic, Cat no.UFC30LG25) by centrifugation at 10000 x G for 5 min.The filtrate extract was collected and the retained pellet of FLs-MS-Ac was washed with 4 x 0.5 mL of IAT, by spinning at 10000 x G for 5 min for each wash.
Analysis of pellets and washes from both methods (Fig. S4): After the final wash the pellet was dissolved in 1.00 mL of 1N NaOH for 1 h.The supernatant (or filtrate) extracts and wash samples were diluted 1:1 with 1N NaOH and kept for 1 h.Next all samples were diluted 1:10 with 1N NaOH then a portion (0.200 mL) was transferred to a well of a black microtiter plate.Fluorescence was measured using a plate reader as described in the general methods.S2.After injection the implants were marked with a permanent marker (Sharpie) by drawing a 12 mm circle around the implant.After injection #5 the animals were euthanized with CO2 then injection #6 was made immediately prior to harvesting tissue.Tissue was harvested to a depth down to the peritoneal membrane, using a 12 mm biopsy punch (Acuderm Acu.Punch P1250).Tissue samples were placed into 2 mL screw cap vials and stored at -80 o C prior to analysis.Untreated tissue samples (X) were also collected to serve as background samples.
Table S2.Dosing schedule and injection map for implantation of RP/FLS-MS mixture into rats.

Determination of FLS and RP in tissue
Tissue samples were thawed then placed into 5 mL snap-cap round bottom centrifuge tubes.Tissue samples were covered with 2.00 mL of IAT buffer, homogenized at 8000 RPM for 1 minute (IKA T25 Ultra-Turrax homogenizer) then allowed to stand for 20 min prior to centrifugation at 5000 x G for 20 min.The supernatant was transferred to a 5 mL Eppendorf tube.The pellet was treated with an additional 2.00 mL of IAT, mixed, allowed to stand for 20 min, then centrifuged at 5000 x G for 20 min.The supernatant was removed and combine with the first supernatant.The pellet was digested in 4.0 mL of 1 N NaOH for 1 h at room temperature.The pellet digest and supernatant (extract) were diluted to exactly 5.0 mL with water.For FLS detection samples were further diluted 5 fold with water and for RP detection samples were measured as is.Samples were added to a black 96-well microtiter plate (0.250 mL/well) and analyzed for fluorescence as described in the general methods (Fig. S5).
Figure S5.Plot of free RP and FLs-MS remaining in injection sites following SC injection of a mixture of RP and FLS-MS-Ac into rats.A) Raw data.B) Normalized data using the ratiometric method.

Tumor xenografts in mice and intra-tumoral injection of FLs-MS~RPR
Female Balb/c mice weighing 21 ± 1.2 g were implanted with 1x10 5 CT26 Murine Colon Carcinoma cells, sc, in 0.100 mL of serum free media.One implant per animal was made.
Tumors were allowed to grow to 251 ± 90 mm 3 (12 days) as determined by measurement with calipers prior injections.Intra-tumoral injection of FLs-MS~RPR slurries (2A, 2B and 2C) were made from 0.250 mL glass syringes with 30 g needles described above (Fig. S3).The 0.025 mL payload of each syringe was injected into a single tumor in two portions (~0.0125 mL each) on either side of the tumor.IAT buffer was injected in the same manner to provide background samples.
Tissue harvesting: At the desired time (hours) after injections, animals were euthanized by cervical dislocation, then the tumors and surrounding tissue were harvested using a 12 mm biopsy punch (Acuderm Acu-Punch P1050).Tissue samples were placed into 10 mL round bottom snap cap centrifuge tubes (Green Bio Research MC01100) containing 4 mL of pH 5.0 IAT buffer.Samples were stored at 4 o C for 40 min prior to being frozen on dry ice then stored at -80 o C for 24 h prior to assay.

Determination of FLS and RP in excised tumors
Tissue samples were thawed to room temperature and immediately homogenized at 8000 RPM for 1 minute (IKA T25 Ultra-Turrax homogenizer).Samples were then centrifuged at 5000 x G for 30 min.The supernatant extract (~4 mL) was transferred to a 15 mL Falcon tube.The pellet washed twice IAT buffer by treating with 2.00 mL of buffer, allowed to rock for 45 min, then centrifuging at 5000 x G for 20 min.The wash extracts were removed and combined with the original supernatant extract to give (~8 mL) of total extract.Extract solutions were treated with 1 mL of NaOH (1 N) and diluted to 10 mL with water.The pellets were treated with 1 mL of 1 N NaOH and then diluted to 10 mL with water.Standards were made by injecting 0.025 mL of each material (2A, 2B and 2C) from dosing syringes into 10 mL of NaOH (0.1 N).After standing for 1 hour the samples were centrifuged at 5000 x G for 20 min, then 0.150 mL of each was loaded into a black 96 well microtiter plate and scanned for fluorescence as described in the general methods.(5).The sterilized amino-MSs were transferred to a 500 mL conical tube and washed with 2 x 300 mL of sterile aqueous 25% acetonitrile (%v/v), separating the amino-MS from the excess solvent by centrifugation (3700 x G, 10 min) between washes.Prior to washing the acetonitrile was sterilized by filtration using a 0.2 µm PTFE 50 mm Disc filter (Saint Gobain PureFlo, D50CF0201N1N-1), whereas the water was sterilized in the autoclave.The amino-MS were then washed with 5 x 200 mL of acetonitrile as described above to give packed amino-MS slurry in acetonitrile (containing 141 μmol amine).This slurry was then treated with triethylamine (4 eq., 565 μmol) and BCN-HSC (1.2 eq., 170 μmol) in 8 mL acetonitrile.The reaction was mixed end-over-end at ambient temperature for 2 h. at which time a qualitative TNBS test ( 5) confirmed that the amines had been consumed.Acetic anhydride (1 eq, 141 μmol) was then added for 0.5 h to cap any residual free amines, and the slurry was washed with 6 x 250 mL acetonitrile as described above.The final packed slurry of BCN-microspheres was 50 mL and contained 141 μmol BCN and was used as is for loading with N3-linker(SO2Me)-SN-38 (below).
The concentration and loading efficiency of the MS~SN-38 was determined by dissolving 20 μL of the packed slurry (20 mg) in 50 mM NaOH (80 μL) for 1 hour at room temperature.The SN-38 content was determined by absorbance of the SN-38 anion by A414.The PEG content in the solution was determined using a previously described colorimetric PEG assay (5).The percent loading of the microsphere slurry was determined to be 98%, by the ratio of SN-38 to PEG (found 196 +/-10 nmol SN-38 / mg PEG, 200 theory).

In vitro release and dissolution of MS~SN-38
Samples of MS~SN-38 (60 μL) were placed in custom made dissolution cells with permeable nylon mesh membranes (Fig. S6).These cells are simply smaller scale versions of previously described dissolution cells (5).The cells were place in 5 mL tubes containing 4.1 mL of 100 mM Na borate buffer pH 9.4 pre-equilibrated to 37°C in a temperature-controlled stirred heat block (Fig. S6).A liquid handling robot ( 5) was programmed to remove 60 μL samples from the reaction buffer at t=0 and specified time points over 50 h.The concentration of SN-38 in the reaction supernatant was determined by absorbance at 414 nm based on previously described methods (5) The A414 of the supernatant vs time were fit to a single exponential using Prism software to determine the release t1/2 for SN-38 was 1.6 h.Similarly, the PEG content of the supernatant was determined using a previously described PEG assay (5) to determine the tRG = 25 h (Fig. S7).

Dose administration and recovery
On Day 0, seven male SD rats were weighed, anesthetized with isoflurane, and their backs shaved.Then, 100 µL of microsphere slurry (containing the mixture of MS~SN-38 and FLs-MS-Ac described above) was injected SC into 4 sites equally spaced on their backs.The location of each injection was marked with circular tattoo ~12 mm in diameter prior to injection.On days 0, 1, 3, 6, 12, 18 and 24, one rat/day was euthanized with CO2, then the hair was removed by shaving and treatment with Nair (Church & Dwight), followed by cleaning with 70% isopropyl alcohol.Using the tattoos for reference, the tissue surrounding the SC injection site was excised down to the peritoneal membrane using a 12 mm diameter biopsy punch (Acuderm Acu-Punch P1250).For each rat, a control skin sample (no injection) was obtained from the flank at least 2 cm away from all injection sites.Each tissue sample was placed into a 1.5 mL Eppendorf tube and stored at -80 °C until analysis.

Analysis of free SN-38 and MS-bound SN-38 at injection sites
To assess free SN-38 present at the injection site tissue samples were thawed in 1 mL 0.5% HOAc, then mixed with a Dounce homogenizer.The pestle was rinsed with 1 mL of 0.5% HOAc which was combined with the homogenized tissue sample.Samples were then clarified by centrifugation (15 min) and the supernatant was removed.The tissue sample was washed with 2 mL 0.5% HOAc, clarified by centrifugation (15 min) and the wash was combined with the previous supernatant to give ~4 mL 0.5% HOAc extract for analysis.A sample of each extract (0.050 mL) was analyzed by HPLC (below) to determine free SN-38 present in the tissue surrounding the injection site.
To assess the content of MS~SN-38 and MS-FLs-Ac at the injection site, the tissue pellets from the HOAc extracts (above) were treated with 0.25 M NaOH (4 mL) vortexed to mix then heated to 80 °C for 1 h to dissolve MS and release MS bound SN-38.The mixture was clarified by centrifugation (15 min), and the supernatant was removed for analysis.A sample of each supernatant (0.050 mL) was analyzed by HPLC to quantify SN-38 and FLs.HPLC analysis.The acid and base extracts of tissue samples described above were analyzed by reverse phase HPLC using the following method: 0% ACN 0-1min, 0-100% 1-30 min, 100% 30-31min, 100-0, 31-32min, 0% 32-33 min all with 0.1% TFA.Fluorescence detection: EX 360nm EM 545nm (0-14.5 min) for SN-38, and EX 442nm EM 520nm, (14.5-30 min) for Fluorescein.SN-38 eluted at 13.2 min, and PEG-fluorescein eluted at 16.7 min.An SN-38 standard curve was generated using 50 µL 0.72 nM to 11.3 uM SN-38 quantitated by HPLC.The fluorescent SN-38 peak area of 50 μL injections was plotted vs. concentration to obtain the standard curve.

Animal Welfare Statement
All animal handling and care was performed by MuriGenics (Vallejo, CA).All experiments were performed using protocol and conditions that conformed to their Institutional Animal Care and Use Committee recommendations.

Figure S3 . 4 )
Figure S3.Dosing syringes and dosing accuracy.A) Dosing syringes containing 0.025 mL single doses of materials 2A, 2B and 2C with purged and capped needles.B) Average

Figure S4 .
Figure S4.Separation of FLS-MSs (A) from free RP (B) after blending with a tissue homogenizer.

Figure S6 .
Figure S6.Small scale PEEK dissolution cell.A) Cell in 5 mL conical tube with stirrer.(B) Top view of the assembly shown in part A. C) Exploded view of cell showing body, caps and nylon mesh membranes.D) 12 position stirring heat block for operation of cells on a liquid handling robot.

Table S1 .
Accuracy of dispensing 0.025 mL of MSs from a 250 μL syringe with a 30 g needle