Cancer is both a local and systemic disease. The mainstay of treatment of many metastatic solid malignancies has been regional, i.e., surgery, radiation, ablation, with or without systemic anticancer agents given intravenously (IV) or orally [1
]. Only small amounts of systemically administered anti-cancer drugs reach the vascular areas of the tumor with less drug reaching cancer cells in a tumor’s hypoxic regions [2
]. The result of systemic dosing is low mass transfer into cancer cells, potentially incomplete dispersion throughout the tumors and poor patient outcomes. These challenges are more pronounced in larger tumors and metastatic disease. Additionally, certain genetic factors of cancer cells, such as those that limit the expression of internalizing drug-transport receptors or that increase efflux pump drug removal, can further reduce intracellular drug concentrations [3
]. These factors contribute to the limited efficacy of systemically administered treatments [4
]. Systemic delivery also distributes a drug throughout the body, which often results in off-target toxicities [5
Direct intratumoral (IT) drug therapy, which has been investigated over the past several decades [8
], could theoretically allow for higher doses to reach the tumor without an increase in systemic toxicity. IT delivery was initially studied with chemotherapeutic agents [9
] using formulations that attempted to enhance residence time in the tumor by the addition of gels, vasoconstrictors, and other retention agents [9
]. However, these approaches failed to improve efficacy over systemic delivery. The disappointing performance of IT dosing historically may be due, in part, to poor drug dispersion throughout the tumor when dosed [18
], a low transmembrane flux rate inherent with receptor transport [3
], and the inability of localized therapies to address the systemic nature of cancer (i.e., the circulating tumor cells and unseen micrometastases) [1
]. Few regional chemical-based treatments have therefore become the standard of care, although transarterial chemoembolization or the use of alcohol dosed into localized small liver tumors is used regularly.
More recently, treatment approaches have shifted from killing the cancer cell to stimulating immune cells. This shift to immune-oncology (IO) treatment has reopened the investigations into intratumoral approaches focusing on activating local immune response. Indeed, a novel genetically modified oncolytic viral-based immunotherapeutic, talimogene laherparepvec (T-Vec), has been approved for IT use [21
] in cutaneous melanoma. The objective of this viral approach is to transfect the granulocyte-macrophage colony-stimulating factor gene into the tumor microenvironment to recruit a local inflammatory response that would promote a systemic immune response. While T-Vec is approved solely for local treatment of localized cutaneous melanoma, the drug has not been shown to improve overall survival or have any effect on distal metastases [22
]. Other local treatment approaches also attempt to recruit the immune system cells into the local tumor microenvironment. Data on several other intratumorally-delivered agents such as STING agonists, RIG-1, and TLR9 have been presented at major cancer conferences [23
]. When the immune system is engaged, even at the single tumor level, there is the potential that a local IT approach could extend beyond the treated tumor. Accordingly, there remains a continued unmet need for the development of direct IT therapies for solid tumors that provide high local efficacy coupled with nontoxic systemic effects.
An ideal formulation for IT delivery would allow the anticancer agent to freely disperse the drug throughout the entire tumor, preferentially enter cancer cells, sparing healthy normal cells, and reaching the intended target either on the cancer cell surface or inside the cell. Amphiphilic molecules are compounds that are soluble in lipids and water systems. Some of these amphiphilic molecules can also bind noncovalently to active drugs, thereby forming supramolecular complexes [24
]. The agents have been used to facilitate protein absorption in the gut for systemic delivery following oral administration [24
]. Such amphiphilic compounds act as delivery agents for the drugs into cells by a passive diffusion process [24
]. These agents have properties that improve the aqueous and lipid solubility of therapeutics to enhance transmembrane cell penetration via a diffusion process without damage to the membrane [24
Herein, we report experimental data using a novel formulation consisting of an amphiphilic compound, 8-((2-hydroxybenzoyl)amino)octanoate (also referred to as SHAO), co-formulated in water at a fixed ratio with two potent agents, cisplatin and vinblastine sulfate. The formulation being referred to as INT230-6. Evidence is described of tissue dispersion, improved tumor kill and attraction of cells for potential immune activation. The new formulation is superior at tumor regression and increasing survival compared to the same drugs at the same concentrations alone given IV or IT without the amphiphilic agent. INT230-6 is now being evaluated as a treatment for solid cancers at clinical sites in the United States and Canada (clinicaltrials.gov
A desirable feature of INT230-6 is selectivity penetrating cancer cells over healthy ones. For a number of tissues, the cancer cell has substantially increased membrane fluidity compared to healthy cells [18
]. An amphiphilic formulation such as INT230-6 with greater lipid solubility would have greater diffusivity through the more fluid cancer cell’s lipid bilayer vis-a-vis the same organ’s healthy cell membrane. Thus, the amphiphilic formulation lipid soluble INT230-6 being physically injected into a tumor has cancer-cell-targeting potential. Toxicology studies conducted as part of the regulatory development process indeed demonstrated that the potent agents in INT230-6 following injection to a normal liver do not harm hepatocytes. SHAO is mixed with and non-covalently bound to its target payloads [24
], once such formulation is in a dilutive environment (i.e., in the blood or the water compartment of cells), the compound is diluted away from the active agent. IT dosing thus distributes the drug payload either into the desired cancer cell or systemically at subtoxic concentrations (depending on amount dosed) via absorption in the vascular compartment.
The anticancer activity of INT230-6 as described below appears to be related to initial cytoreduction, recruitment of dendritic cells to process antigens created from the tumor that lead to a reduction in non-treated tumor volumes. In addition, immunological response is observed to further eradicate cancer cells and protect the animals from re-challenge.
Direct IT therapy has been proposed as a means of improving the therapeutic index of active IV formulated drugs [9
]. Unfortunately, the direct IT therapeutic strategies investigated to date have had limited success [8
], primarily due to the lack of efficacy beyond the area injected. It is postulated that the lack of efficacy of IT approaches has been due to poor drug dispersion within the tumor, lack of absorption by the tumor coupled with ejection post injection. In addition, a limited penetration into the cancer cell, as well as the inability to affect distal sites of disease have also decreased the utility of past or current IT approaches. Through the use of a novel formulation chemistry, the research conducted in our studies has identified a potent tumor tissue dispersing formulation containing a fixed ratio of an amphiphilic cell penetration enhancer molecule, SHAO, formulated with cisplatin and vinblastine and designated as INT230-6. Improved drug absorption/dispersion using INT230-6 led to significantly greater tumor regression and increased survival benefit compared to the drugs alone given IV or IT. Using large tumors, INT230-6 results in a number of CRs in multiple tumor types. Furthermore, the results show that INT230-6 treatment increased tumor necrosis and the appearance of certain immune cell infiltrates in the tumor microenvironment. Indeed, the presence of tumor antigens released by the dying cells of a highly necrotic tumor appears to create an environment favorable to induce a systemic immune response. This hypothesis is supported by the re-challenge experiment showing that complete responders from previous INT230-6 treatments developed immunity against the same cell lines. Literature studies show that cisplatin has activity in multiple tumor types [35
], and platinum agents have the ability to induce immunologic cell death in part by release of calreticulin [36
]. Another possibility could be that the enhancer-based IT delivery modality may increase the calreticulin release, located in storage compartments associated with the endoplasmic reticulum, to the cell surface. Cisplatin can also result in high mobility group box 1 (HMGB1) protein production, thereby stimulating mature dendritic cell processing through interaction with toll-like receptor-4 (TLR-4) [36
While repeat INT230-6 (cytotoxic) treatment to the same tumor in theory could impair or kill the beneficial immune cells recruited to the tumor microenvironment, the data generated indicate that repeated INT230-6 treatment yielded superior CR rates and overall survival compared to single treatment. This finding leads to a preferred clinical regimen of repeated INT230-6 cycles.
The current clinical success of checkpoint antibodies in many major tumor types is limited due to their inability to overcome T-cell suppression or improve antigen recognition. The combination of INT230-6 to attenuate tumors and improve antigen presentation for immune recognition given IT either alone or with checkpoint inhibitors may have promise for improved clinical anticancer activity. There is further experimental evidence in mice that the systemic administration of chemotherapeutic agents impairs the immune response to a PD-1 antibody, while the local administration of these potent agents potentiates immune activity [38
]. In the clinic, Ariyan [39
] reported that the isolated limb infusion of melphalan with systemic ipilimumab, an CTLA-4 antibody, in patients with in transit melanoma, showed a durable increase in efficacy over ipilimumab alone.
INT230-6 is currently undergoing phase 1/2 dose escalation in a clinical trial investigating repeating doses, dose frequency and drug load per tumor alone and in combination with a commercial PD-1 antibody (pembrolizumab) and a CTLA-4 antibody (ipilimumab) in several different refractory solid tumor cancers (NCT03058289). Patients will be followed for safety and will be evaluated for both injected and bystander tumor responses. The study also tracks the pharmacokinetics of each INT230-6 component. In addition to clinical endpoints, blood and tumor samples will be evaluated to look at the tumor microenvironment and central immune compartment. A better understanding of the tumor and its stroma will enable the customization of IT-designed formulations such as INT230-6 for improved patient outcomes.
4. Materials and Methods
Experiments included the evaluation of INT230-6 dispersion in tumors with and without the enhancers; the effect in vitro on cancer cell morphology; the evaluation of certain immunomodulatory effects of IT treatments, including immune cell infiltration into the tumor microenvironment; the growth inhibition, tumor regression and synergy of INT230-6 when combined with checkpoint inhibitors.
The penetration enhancer molecule in INT230-6 is a 279 molecular weight agent, 8-((2-hydroxybenzoyl)amino)octanoate (referred to as SHAO in solution and obtained from AMRI Global Albany, New York USA), and is considered an excipient with no pharmacologic activity of its own. The compound’s molecular formula is C15
and the structure is shown in Figure 8
. The two active pharmaceutical ingredients in the INT230-6 formulation are cisplatin (CAS ID15663-27-1) and vinblastine sulfate (CAS ID1143-67-9) both obtained from Tocris Bioscience a division of Bio-Techne Corporation, Minneapolis, MN USA. The composition of INT230-6 is 10 mg/mL of SHAO, 0.5 mg/mL of cisplatin, and 0.1 mg/mL of vinblastine. To prepare INT230-6, the desired amount of SHAO was dissolved in a ~0.35 M NaOH solution, followed by the addition of 0.1% Tween80 (Cat No. AC278632500, Fisher scientific, Waltham, MA USA) and the required amount of the active agents, and then filtered for sterilized dosing.
4.2. Cell Lines
Murine cell lines included 4T1 breast cancer cells obtained from the American Type Culture Collection (Manassas, VA), human BxPc3-luc2 pancreatic adenocarcinoma cells obtained from Caliper Life Sciences (Hopkinton, MA USA), and Colon-26 murine adenocarcinoma cells obtained from the National Cancer Institute, Bethesda, MD USA.
All experiments were approved by the US Public Health Service Policy on Humane Care and Use of Laboratory Animals and carried out either at the MI Bioresearch division of Covance / LabCorp in Ann Arbor Michigan USA, CR Discovery Services, Morrisville, NC USA (a division of Charles River Laboratories (CRL), or Taconic Biosciences Hudson, NY USA. Female mice (BALB/c or Nude severely compromised immune deficient mice, obtained from Charles River Laboratories Wilmington, MA, USA, aged 6 to 12 weeks were used. MI BioResearch, Taconic Biosciences and CR Discovery Services specifically comply with the recommendations of the Guide for Care and Use of Laboratory Animals with respect to restraint, husbandry, surgical procedures, feed and fluid regulation, and veterinary care (Supplementary Materials Text S1
Animal health and behavior were monitored twice per day. Any individual animal with a single observation of >30% body weight loss or three consecutive measurements of >25% body weight loss was euthanized.
The endpoint of the experiment was a tumor volume of 2000 mm3. When the endpoint was reached, the animals were to be euthanized immediately.
The animals were terminated by CO2 asphyxiation or cardiac puncture depending on the lab and study. This was followed by cervical dislocation. All studies were approved by the specific site’s Institutional Animal Care and Use Committee (IACUC). Experiments were performed at multiple locations (Supplementary Materials Text S2
4.4. Drug/Enhancer Dispersion Studies in Murine Models
Drug dispersion in tumors was assessed in three studies at three laboratories in female nude mice bearing BxPc3-luc2 pancreatic tumors. In one experiment, mice approximately 5 to 6 weeks of age were implanted with cryopreserved BxPC3 fragments (MI3319) The site’s Institutional Animal Care and Use Committee (IACUC) approvals were approximately March 2018 and July 2018. When tumors reach ~750 mm3, the fragments were passed to additional animals and then those additional animals were passed fragments until a sufficient number of animals had tumors on their right flank of sufficient size to conduct the study. Three drops (~150 µL) of India Ink was added to a 10-mL vial of INT230-6 clinical supplies grade drug product. A control solution of 10 mL using the same aqueous vehicle with 0.1% Tween80, cisplatin (0.5 mg/mL) and 150 µL of India Ink was also prepared. Eight animals were dosed with INT230-6/Ink solution and six animals were dosed with drug/Ink alone. All animals were oriented in the same position for dosing. A 26-gauge needle with a butterfly valve was placed in the center of the tumor at the same angle for each of the injections. IT dosing was performed over 90 s using a dosing pump approximately 0.75 cm from the tumor surface. A drug volume of 0.075 or 0.225 mL was administered at tumor volume ratios of approximately 1:11 and 1:4. Animals were sacrificed, necropsied and examined for solution in the abdominal and chest cavities. All tumors were excised, split in half along the same axis and examined immediately following dose completion. Direct measurement and observations on Ink containing solution diffusion diameters and ex-tumor leakage were made.
4.5. Tumor Growth Inhibition and Survival
Growth inhibition studies assessed the pharmacodynamic effects of INT230-6 in in vivo murine models. The effect of INT230-6 on tumor growth inhibition and survival was assessed in mice bearing large tumors in the hind flank compared to controls (i.e., no treatment, enhancer alone, or drugs alone dosed IV and/or IT). Abnormal behavior (e.g., food consumption, body weight, activity levels) was assessed as a measure of toxicity. Tumor volume was calculated by caliper using the measured width squared (w2) in millimeters (mm) multiplied by the length (mm) divided by 2. The volume (V) equation was V = (w2 × L)/2. Tumor size and animal survival were assessed over time, and animals were terminated per protocol using the methods described in S2 Text once the tumor volume was above 2000 mm3. A complete response (CR) was defined as tumors completely disappearing with no measurable caliper areas.
4.6. Colon-26 Studies
Colon-26 was chosen as the primary cell line for the first set of growth studies as this murine cancer type is commonly used to test agents in syngeneic animals. Larger initial tumor volumes represent advanced disease and are a much more challenging model to demonstrate tumor regression or increased survival. In general, for Colon-26 studies, BALB/c female mice were injected in their right flank with 1 × 106 Colon-26 tumor cells (cell injection volume, 0.1 mL/mouse). Animals were treated when the mean tumor volume reached 300 mm3 after randomization and only mice with tumors smaller than 500 mm3 were enrolled in the study. The mean standard deviation of the groups was approximately 16 to 17 mm3. Randomization typically occurred 17 to 19 days following cancer cell inoculation. Mice receiving IT treatment were administered 0.1 mL per 400 mm3 tumor volume for 3 to 5 days depending on the study. IT and IV drug alone doses were comparable in total drug dosed. Experiments in the series were Colon E230 and E236 with IACUC approvals made approximately on August 2013 and December 2013.
4.7. 4T1 Studies
To confirm the anti-cancer activity of INT230-6, growth inhibition studies were conducted in a second cell line, 4T1. Due to the rapid formation of metastases, mean tumor volumes for 4T1 studies were approximately 125 mm3 after animal randomization into test groups. BALB/c female mice were injected in their right flank with 1 × 106 cells. The cell implant volume was 50 µL and randomization occurred 7 to 9 days following cancer cell inoculation. Mice receiving IT treatment were administered 0.1 mL per 400 mm3 tumor volume for 3 to 5 days depending on the study. In one study, INT230-6 was dosed once daily (QD) for three days (QDx3) at 0.1 mL IT, which contained at total 50 µg of cisplatin and 10 µg of vinblastine. Anti-mPD1 (RPM1-14, cat# 5792-599016j1, BioXCell, Lebanon, NH USA) antibodies were dosed in two cycles of three consecutive days and 3 days off (Q3Dx2; 3 off) x2. The experiment reported in 2.5 was MI2477 with IACUC approval made approximately on January 2016.
4.8. In Vitro Cell Membrane Retention Studies
To determine whether the enhancer does not disrupt to the cancer cell membrane, in vitro studies were conducted investigating SHAO alone in a static system. Solutions of the enhancer at various concentrations together with a control were incubated for up to 24 h with the Colon-26 cell line. Photomicrographs showing the morphology of the cells were taken.
4.9. Checkpoint Combination Studies
There is potential for an increased response of checkpoint inhibitors using INT230-6 due to the likely presentation of antigen and recruitment of T-cells following cell death. Studies using IT dosed INT230-6 concurrently or sequentially with IV administered monoclonal antibodies against cytotoxic T-lymphocyte antigen 4 (CTLA-4) (9H10, lot# 5294/0814, BioXCell, Lebanon, NH USA) and/or programmed death-1 (PD-1) (RPM1-14, lot# 5311-4/0714C, BioXCell, Lebanon, NH USA) were conducted. BALB/c female mice aged 8 to 12 weeks were injected in their right flank with 1 × 106 Colon-26 tumor cells using a cell injection volume of 0.1 mL/mouse at Charles River Laboratories (CRL). INT230-6 was administered IT to mice with tumors approximately 300 mm3 in size; CTLA-4 and PD-1 antibodies were administered IV. In one study, the following groups were assessed: no treatment (Group 1; control); INT230-6 administered IT once daily for 5 days (Group 2); INT230-6 administered IT once daily for 5 days (with 9 days off) for 3-dose cycles (Group 3); INT230-6 administered IT once daily for 5 days and PD-1 antibody administered twice weekly for 2 weeks beginning on Day 0 (Group 4); INT230-6 administered IT once daily for 5 days and PD-1 antibody administered twice weekly for 2 weeks beginning on Day 10 (Group 5); INT230-6 administered once daily for 5 days and CTLA-4 antibody administered on Days 10, 13, and 16 (Group 6); INT230-6 administered once daily for 5 days, PD-1 antibody twice weekly for 2 weeks beginning on Day 10, and CTLA-4 antibody on Days 10, 13, and 16 (Group 7); and PD-1 antibody administered twice weekly for 2 weeks beginning on Day 10 and CTLA-4 antibody administered on Days 10, 13, and 16 (Group 8; control). Mice were observed frequently for health and overt signs of any adverse treatment related side effects and noteworthy clinical observations. Considering the length of the study (100 days), the mice occasionally found dead in the cages and not reaching 2000 mm3 were removed from the study for a better representation of the tumor growth curves. The list of removed mice is as follows: Group 1 (1 animal), Group 2 (3 animals), Group 3 (2 animals), Group 4 (1 animal), Group 5 (1 animal), Group 6 (1 animal), Group 7 (1 animal), Group 8 (1 animal).
4.10. Immune Cell Flux Immunohistochemistry Evaluations
The tumors from two groups of 12 animals (INT230-6 treated and untreated) were evaluated histologically for immune cell markers (CD4, F4/80, CD11c, and CD335). Each animal was inoculated with 1 × 106 Colon-26 colon cancer cells on one flank. When group tumors reached a mean value of 323 mm3 per group, treated animals received daily IT injections of INT230-6 for five consecutive days. At six timepoints over a period of 10 days, two animals from each group had their tumors excised, the amount of necrosis estimated, and the tissue fixed and shipped to CRL (Fredericksburg, MD, USA) for immunohistochemical evaluation. Immunohistochemical staining was performed to detect various immune cell markers (CD4, F4/80, CD11c, and CD335) to evaluate the presence of infiltrating mononuclear cell populations within the tumor. To detect the markers within the test tissues, the detection antibodies were applied to acetone/formalin-fixed (CD335) or formalin fixed (CD4, F4/80, and CD11c) cryosections of the tumors. Blocking buffer composition was PBS + 1% bovine serum albumin (BSA); 0.5% casein; and 1.5% normal donkey serum. Following the protein block, the primary antibodies (Rat anti-mouse CD4 (cat# 553043 BD Pharmingen, Woburn, MA USA), Rat anti-mouse F4/80 ( cat# MCA497R, BioRad Kidlington, UK formerly AbD Serotec) and Rat IgG2a, kappa (cat# 559073, BD Pharmingen, Woburn, MA USA) (1 μg/mL for 1 h), or none (buffer alone as the assay control)) were applied to the slides at a concentration of 1 μg/mL for one hour. Hamster anti-mouse CD11c (cat# MCA1369GA, AbD Serotec, Hercules, CA USA), Hamster IgG (cat# 007-000-003, Jackson ImmunoResearch West Grove, PA USA), or none (buffer alone as the assay control) was applied to the slides at a concentration of 1 μg/mL for one hour. Rat anti-mouse CD355 (cat# 137602, BioLegend Dedham, MA USA), Rat IgG2a, kappa (cat# 559073, BD Pharmingen, Woburn, MA USA), or none (buffer alone as the assay control) was applied to the slides at a concentration of 10 μg/mL for one hour.
Experiments in the series for sections 4.09 and 4.10 were CR-E262, CRL-E241, IACUC approvals made approximately November 2014 and July 2014.