Challenges and Key Considerations of the Enhanced Permeability and Retention Effect for Nanomedicine Drug Delivery in Oncology

The field of nanomedicine, despite being conceptualized as far back as the 1980s, is only now transitioning in a broad sense from academic research to drug development and commercialization. In oncology, unique structural features of many solid tumors, including hypervasculature, defective vascular architecture, and impaired lymphatic drainage leading to the well-characterized enhanced permeability and retention (EPR; ref. 1) effect, are key factors in advancing this platform technology. However, the EPR effect has been measured mostly, if not exclusively, in implanted tumors with limited data on EPR in metastatic lesions. Dextran-coated iron oxide nanoparticles (25–50 nm) have been used clinically for several years (2) to measure permeability and retention noninvasively by MRI (3). Furthermore, tumor response alone is no longer considered a good endpoint, at least from the health authority point of view. This is exemplified by the recent U.S. Food and Drug Administration (FDA) withdrawal of bevacizumab (Avastin) for patients with metastatic breast cancer where impressive tumor responses were seen but bevacizumab showed no improvement in overall survival. Thus, limitations and challenges both in understanding tumor structural features and correlating them with the technology must be addressed and additional critical data need to be generated before nanotechnology-based drug delivery approaches can be fully realized in clinical use in patients with cancer. A one-day workshop was convened at the NIH on October 10, 2012, to specifically address key issues related to understanding of EPR effect and its use to achieve the maximum therapeutic effect with drugs using nanoparticle carriers.

This workshop was organized by the Alliance for Nanotechnology in Cancer and its recently formed public–private partnership consortium, TONIC (Translation of Nanotechnology in Cancer), in response to several questions raised by industry members of TONIC. The main purpose of this meeting was to gain better understanding of the EPR characteristics impacting the use of nanoparticles in the clinic. Experimental evidence of EPR in animal models and humans, clinical relevance of EPR, gaps in knowledge, and ways to address these gaps were all discussed.

The workshop composed of 8 talks covering topics ranging from methods to investigate EPR in preclinical and clinical studies including diagnostic imaging, to the ramifications of EPR for enhanced drug uptake by different tumors and the predictability of preclinical and clinical outcomes. The session opened with an overview of the nanotechnology programs in cancer, funded by the Alliance for Nanotechnology in Cancer (NCI), and was followed by an introduction to TONIC, a corporate partnership model of the public, private, and academic sectors, to accelerate the translation and development of nanotechnology solutions for the early detection, diagnosis, and treatment of cancer. This was followed by scientific presentations relating to the key questions identified at previous TONIC meetings. The discussions at the workshop focused on two key themes, namely, heterogeneity of EPR in tumors and factors that influence EPR effect.

EPR exists in tumors and can be exploited for selective delivery of drugs to tumor by nanotechnology. However, there is significant heterogeneity within and between tumor types. It was noted that different tumor types have different pore dimensions in the vasculature and that the maximum pore size changes with the location for a given type of tumor (i.e., primary vs. metastases). In addition, there may be differences in vessel structure within a single tumor type. Thus, to understand whether a tumor is likely to respond to a nanoparticle-based drug that relies on EPR for delivery, an image-guided patient selection or diagnostic approach can potentially prove useful to profile and select tumor types and patients with tumors conducive to such delivery. Hiroshi Maeda (Sojo University, Kumamoto, Japan), who first proposed the EPR effect over 25 years ago (1), suggested a number of ways one can augment the EPR effect. These included increasing the blood pressure during infusion of a nanomedicine or macromolecular drug using angiotensin-II (e.g., blood pressure increase from 100 → 150 mmHg). Other methods involve vascular mediators such as nitroglycerin, ACE inhibitor, or PGE1 agonist (beraprost) and these have been shown to be effective in in vivo tumor models resulting in better tumor delivery (2- to 3-fold increase), linked to improved therapeutic effect (4).

The following factors influence the EPR effect in tumors: (i) the nature of both the vascular bed and surrounding stroma, the presence or absence of functional lymphatics and interstitial hydraulic conductivity impacting interstitial pressure along with mechanical stresses generated by cancer and stromal cells impacting the extracellular matrix; (ii) tumor size, type, and location (including primary tumor versus metastatic lesions); (iii) extent of macrophage tumor infiltration and the activity of the mononuclear phagocytic system (MPS), which can vary between and within tumor types plus patient characteristics (e.g., age, gender, tumor type, body composition, treatment). These factors lead to accumulation of nanoparticles in both normal tissues and in different sections of the tumor, for example, in the periphery, viable tumor, and necrotic sections; and (iv) co-medications, which may impact, among other things, stroma and blood pressure (hypertension increases tumor blood flow). In addition, several vascular factors (Table 1; ref. 4), such as nitric oxide generators (5) and bradykinin potentiators, that is, ACE inhibitors that lower blood pressure, are known to affect EPR and are relatively safe and inexpensive to combine with a nanoparticle drug (4).

A fundamental limitation in evaluating EPR and the factors that affect EPR is poor understanding of which preclinical tumor models recapitulate patients with solid tumors. The factors affecting delivery of nanoparticles to tumors in preclinical models, such as tumor growth environment, vasculature, functional MPS, etc., appear to vary based on the cancer model [e.g., syngeneic flank xenograft, orthotopic xenograft, genetically engineered mouse model (GEMM)]. Thus, future studies will need to systemically evaluate these factors in preclinical models and in patients with various solid tumors and determine whether the models represent all aspects of the EPR effect.

The observed heterogeneity in EPR may be a contributing factor to the limited impact of nanoparticle-based drugs with reductions in toxicity and gains in overall survival as compared with small-molecule anti-cancer agents. Table 2 summarizes objective data on the survival benefits from nanotherapeutics approved to date. Further understanding and predictability of EPR function in primary tumor and its metastatic sites through the use of imaging studies may aid the development of future, effective nanodrugs. Correlation of EPR activity to clinical responses would likely provide direct clinical data to determine whether tumors with high EPR tumor activity will be more amenable to effective treatment using nanoparticle-based therapies (5). It was noted that the diversity of nanoparticle characteristics and API used is expected to impact the applicability of such correlations across different nanoparticle platforms and products.

The optimal patient selection or diagnostic aid to measure the EPR activity within a patient needs to be further defined. Ideally, this would involve a single imaging agent that is generalizable to all nanoparticles. Given the heterogeneity of nanoparticle-based systems—size, shape, charge characteristics, etc.—a specific diagnostic agent might, however, be required to predict likely response to a particular nanoparticle relying on EPR delivery. The use of contrast agents and MRI to measure the enhanced permeability (EP) component of the EPR effect might be one generic method. Others might include a defined nanoparticle of a fixed size (∼100 nm) labeled with an appropriate imaging agent—for example, Cu⁶⁴ for positron emission tomography (PET) or fluorescent marker for near-infrared fluorescence (NIRF). There is precedence for a range of labeled liposomes and iron oxide–loaded nanoparticles for imaging, but there are very few human clinical studies on nanoparticle imaging that can effectively address the prevalence of EPR. In one such study, the biodistribution and pharmacokinetics of ^[111]In-labeled PEGylated liposomes was evaluated in patients with locally advanced cancers. Positive tumor images were obtained in 15 of 17 studies, although levels of tumor liposome uptake varied between and within tumor types (6).

Eva Sevick-Muraca (The University of Texas Health Science Center, Houston, Texas) discussed the use of NIRF to image lymphatic flow and with fluorescent agents to detect cancers. This technique is light based and the fluorescent dye has no half-life and can be repeatedly excited, making it more appropriate for imaging of nanoparticle accumulation over longer timeframes than radioactive imaging agents with short half-lives (7). While NIRF is considered to be a combination product by the FDA and has a maximum tissue penetration of 3 to 5 cm, such devices are not yet available in hospitals and may not have the right sensitivity at this time to detect the marker agent. The ability to image lymphatic function in the tumor vicinity could also provide a means to assess interstitial pressure imbalances. Efforts are underway to include dual-labeling PET for presurgical imaging and then NIR guidance during surgery (8). It is anticipated that PET will remain a crucial tool for clinical imaging and that the optical imaging counterpart will add value rather than being a replacement.

Ways to enhance the EPR effect in tumors were discussed and included drugs that impacted the vasculature (4)–for example, VEGF-based antagonists leading to vessel normalization, agents causing hypertension and increasing tumor blood flow, and agents that modulate the tumor matrix. Agents that generate nitric oxide [nitroglycerine or ISDN (isosorbide dinitrate)] were also shown to be effective in humans (4, 5). ACE inhibitor (e.g., enarapril), which potentiates the action of bradykinin, is also effective (4). Further work is required to validate the benefits of such agents in the context of exploiting the enhancement of EPR effect in the clinical setting (4, 5). It was suggested that both optimization of the nanoparticle and optimization of the tumor microenvironment were required for optimal delivery. Rakesh K. Jain (Harvard Medical School, Boston, Massachusetts); hypothesized that normalizing the vasculature, extracellular matrix, and lymphatics will lead to better delivery of drugs (9). However, normalized vasculature means that the average pore is smaller and this may require the use of smaller nanoparticles (∼20 nm particle size). Overall, the biologic impact of the abovementioned vascular effectors on delivery of nanoparticles of varying composition, shape, and flexibility needs significant further work.

The role of the lymphatics in tumor biology and nanoparticle delivery was discussed. This highlighted the need to consider changes in physiologic status, both in the acute and in long-term functionality of lymphatics in patients with cancer influenced by inflammation, tumor burden, or treatment. This is an area of active research and imaging techniques are being developed that will allow this to be explored in more detail.

In terms of animal tumor models to evaluate the EPR effect, subcutaneous flank tumor xenografts were thought to offer limited value. The vasculature of such models often resembles the vasculature found in very high EPR tumors, for example, renal tumors irrespective of tumor type, and thus probably gives a false impression about the benefit of nanoparticle-based drugs relying on the EPR effect in most tumor settings. The workshop participants felt that better options are provided by metastatic, orthotopic, and GEMM-based models, although these need further characterization and validation. Primary tumor explants may be another option to model delivery to tumor types with high stromal content. Further work is required to understand how to use the preclinical tumor models to investigate drugs relying on the EPR effect for activity and to understand how they reflect the heterogeneity seen in clinical disease. The site of the tumor was also considered to be important, and a more systematic assessment of vasculature architecture versus site of tumor was recommended.

Omid Farokhzad (Harvard Medical School) discussed the advantages of including a targeting agent on the nanoparticle to enhance the retention component and/or enable delivery of drug directly into the tumor cell via internalization of the nanoparticle. The majority of the currently available clinical data on nanoparticle oncology drugs relate to passively targeted liposomal drugs. Recently, several actively targeted nanoparticle products have also entered clinical development, including liposomes and polymeric particles containing payloads ranging from conventional cytotoxic drugs to genes expressing tumor suppressors (10). These particles are targeted to various tumor markers including the transferrin receptor HER-2 and prostate-specific membrane antigen (PSMA) using either protein or small-molecule ligands. Recent data were presented for BIND-014 (11), a docetaxel-encapsulated polymeric nanoparticle targeted to PSMA, which is expressed on the surface of prostate cancer cells and nonprostate solid tumor neovasculature. In preclinical studies, BIND-014 increased the concentration of docetaxel in PSMA-expressing solid tumor xenografts by 5- to 10-fold. In a phase I clinical trial in patients with advanced solid tumors, BIND-014 displayed signals of antitumor efficacy in patients with advanced and metastatic cancer at low doses and in tumors where conventional docetaxel has minimal activity. With progress in polymeric nanoparticle engineering, similar approaches are also being applied to existing and developmental anticancer drugs, including other cytotoxics and molecularly targeted agents such as kinase inhibitors, and it will only be a matter of time before these advances will ultimately impact the treatment of cancer.

William Zamboni (University of North Carolina, Chapel Hill, North Carolina) characterized the pharmacologic properties of nanoparticles in vivo as part of preclinical and clinical studies. He stressed the importance of the MPS, tissue distribution, and potential tumor delivery on the clearance of nanoparticles. There is a bidirectional interaction between monocytes and liposomal agents and potentially other nanoparticle agents (12, 13). Monocytes internalize liposomes, which then releases the drug from the liposome and leads to toxic effects to the monocytes. The tissue distribution and tumor delivery of nanoparticles may involve MPS-mediated and non–MPS-mediated mechanisms where uptake of nanoparticles by circulating MPS cells compared with tumoral macrophages may result in different tumor drug exposure and responses. Dr. Zamboni has developed an ex vivo flow cytometry–based, high-throughput screening platform (HTSP) system called PhenoGLO-HTSP to measure the clearance of nanoparticles by the MPS and bidirectional interaction between the MPS and nanoparticles, conjugates, and antibody–drug conjugates. Importantly, this method also predicts nanoparticle pharmacokinetics and pharmacodynamics in humans where the MPS system seems to drive the clearance, efficacy, and toxicity of nanoparticle agents. PhenoGLO-IT can measure MPS function in a blood sample from patients as a method to individualize the dose of nanoparticle agents and/or as a biomarker for predicting pharmacokinetics and pharmacodynamics (response and toxicity) of nanoparticles.

The workshop participants felt that as our understanding of nanoparticle delivery to tumors increases, the emerging nanoformulations should be considered both as a general formulation strategy in drug development and as a selected strategy to improve delivery profiles of existing or failed drugs.

During discussions at the conclusion of the symposium, participants recommended the formation of a working group to establish translational and clinical procedures for integrated clinical trials involving nanotherapeutic constructs and accompanying imaging approaches. Such translational studies and clinical trials would enable further understanding and predictability of EPR function in a tumor and its primary or metastatic sites and may be critical for the development of future effective nanodrugs and predictive of antitumor response (14). An additional recommendation from this workshop was to generate a position paper highlighting key translational studies that should be conducted and parameters that should be monitored in nanoparticle drug delivery clinical trials to enable testing of various hypotheses for effective nanoparticle delivery (tumor perfusion, vascular permeability, interstitial penetration, retention, lymphatic function, MPS activity, blood pressure, fluid and solid stresses, others). In coming months, symposium participants will actively pursue these key recommendations and develop the necessary tools required to advance the scientific translation of the nanotechnology platform in the oncology therapeutic area.

D.C. Blakey has ownership interest (including patents) in AstraZeneca. R.K. Jain is employed (other than primary affiliation; e.g., consulting) as a cofounder and board member of XTuit Pharmaceuticals, and as a board member of Hambrecht & Quist Healthcare Investors and Hambrecht & Quist Life Sciences Investors; has a commercial research grants from MedImmune, Roche, and Dyax; has ownership interest (including patents) in XTUit Pharmaceuticals; and is a consultant/advisory board member of Enlight Biosciences, SynDevRx, Dyax, Noxxon Pharmaceuticals, and Zyngenia. E.M. Sevick-Muraca has ownership interest (including patents) in NIRFImaging, Inc. W. Zamboni has ownership interest (including patents) in PhenoGLO Technologies. O.C. Farokhzad has ownership interest (including patents) in and is a consultant/advisory board member of BIND Bioscience, Selecta Bioscience, and BLEND. A. Gabizon has a commercial research grant from Janssen Pharmaceuticals. No potential conflicts of interest were disclosed by the other authors.

Conception and design: U. Prabhakar, D.C. Blakey, H. Maeda, R.K. Jain, E.M. Sevick-Muraca, O.C. Farokhzad, A. Gabizon, P. Grodzinski

Development of methodology: U. Prabhakar, P. Grodzinski

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): U. Prabhakar, P. Grodzinski

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): U. Prabhakar, H. Maeda, P. Grodzinski

Writing, review, and/or revision of the manuscript: U. Prabhakar, D.C. Blakey, H. Maeda, R.K. Jain, E.M. Sevick-Muraca, W. Zamboni, O.C. Farokhzad, S.T. Barry, A. Gabizon, P. Grodzinski

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): U. Prabhakar, H. Maeda, P. Grodzinski

Study supervision: U. Prabhakar, P. Grodzinski