Radiofrequency ablation (RFA) is now used in clinical practice to treat some malignant tumors, yet it suffers from serious limitations 1234. These shortcomings include: 1) RFA is currently an invasive treatment requiring insertion of needle electrodes directly into the tumor(s) to be treated; 2) incomplete tumor destruction occurs in 5% – 40% of the treated lesions, particularly if lesions are > 4–5 cm in diameter; 3) the treatment is nonspecific with both malignant and normal tissues around the needle electrode undergoing thermal injury; 4) complications arise in up to 10% of patients, frequently related to thermal injury to normal tissues; 5) and invasive RFA is limited to treatment of tumors in only a few organ sites (liver, kidney, breast, lung, bone) 56. Interestingly, the tissue penetration in humans by focused external RF energy fields is known to be excellent 7. In theory, non-invasive RF treatment of malignant tumors at any site in the body should be possible, but such a treatment would require the presence of intracellular or intratumoral agents that release heat under the influence of the RF field. For such a novel RF treatment approach to be effective, it will require identification of agents that have little or no intrinsic cellular or tissue toxicity that can also be targeted or directed to malignant cells while sparing normal cells. Clearly, a non-invasive approach with the potential to treat many types of cancers effectively with minimal or no toxic effects to normal cells would be highly beneficial.

Nanoparticles have piqued the interest of the medical community for use in cancer diagnosis, treatment, and as delivery vectors for biologic or pharmacologic agents 8910111213141516. The ability to affect diagnostic or therapeutic changes on a nanoscale could provide significant gains in medical care. Gold nanoparticles (GNPs) are particularly interesting for several reasons. First, they are easily prepared. Additionally, the binding of molecules to the GNPs in order to target cancer cells, including antibodies, carbohydrates, and pharmacologic agents, is easily achieved 1718192021. Also, the GNPs themselves have anti-angiogenic properties 13.

A previous study revealed that gold-silica nanoshells release significant heat when exposed to near-infrared (NIR) light (650–950 nm) and have been used to produce thermal cytotoxicity in vitro 22. Unfortunately, this treatment approach is mechanistically limited to use in superficial malignant tumors because of the minimal tissue penetration (< 2–3 cm depth) by NIR wavelength light 23. However, the gold-silica nanoshell study demonstrated that nanogold has potential clinical use as a thermal conductor of non-invasive energy sources. Gold, like most metals, is an excellent conductor of electrical and thermal energy, thus we studied the potential role of GNPs as intracellular molecules that would release heat when treated with RF irradiation.

We hypothesized that 1) the addition of GNPs to hepatocellular and pancreatic human cancer cell lines would not be intrinsically cytotoxic to the cells and 2) cancer cells containing GNPs exposed to a focused, non-invasive RF field would develop lethal, thermal-induced injury.

The application of nanomaterials to the biohealth arena is an exciting prospect given that most cellular chemical and enzymatic interactions occur on the nanoscale. Therefore, the ability to manage or modify these processes with engineered molecules represents a new frontier for therapeutics.

Conventional RFA is a useful treatment option for destruction of hepatic malignancies, both primary and metastatic lesions, as well as a few additional solid cancers 4. Currently, RFA is limited in the size of the tumors that it can effectively treat 35. Treating tumors larger than 5 cm in diameter with invasive RFA results in incomplete tumor destruction in 10–40% of cases 245. The tumor must be treated with an ablation needle precisely placed to assure optimal tumor destruction. In treating hepatic malignancies, the RF energy applied to the invasive needle electrode produces indiscriminate heating of any tissue type within which it is placed, including normal liver parenchyma, bile ducts, and other organs or structures in proximity to the malignant cells. Additionally, tumor location can prevent percutaneous or laparoscopic (minimally invasive) approaches for RFA. Theoretically, an external RF field generator would eliminate the need for an invasive needle electrode, be able to focus energy at any tumor location and body site, and not be limited by the size of the tumor. In order to produce thermally-induced cancer cell death in response to the RF field in vivo, intracellular or intratumoral resonant or metallic heat-producing molecules are required. GNPs are excellent conductors of electrical and thermal energy and in our system provide non-specific RF targeting to human gastrointestinal cancer cells in vitro. GNPs appear to be taken into the cancer cells in vitro by endocytosis with evidence of cytoplasmic vesicles containing GNPs seen in our electron microscopy images. We have evidence that solid tumor treatment in vivo is feasible and effective using intracellular single-walled carbon nanotubes as the heating releasing entity 24, but this in vivo approach needs to be validated using GNPs. Ideally, GNPs can be targeted to malignant cells in vivo by attaching tumor-specific or tumor-related targeting molecules such as antibodies, peptides, or pharmacologic agents.

The data here represent the combination of these two novel approaches, intracellular GNPs and a unique non-invasive RF field generator. Other researchers have demonstrated some decreased cellular proliferation with GNP exposure. For example, GNPs have been shown to have anti-proliferative activity in multiple myeloma cells 25. GNPs were not cytotoxic to these myeloma cells and the anti-proliferative activity was reversible. GNPs are not cytotoxic or anti-proliferative in vitro in the two solid tumor cancer cell lines studied here. This is demonstrated in the MTT assays, PI-FACS control specimens without RF, and the normal TEM appearance of GNP-treated Hep3B and Panc-1 cancer cells not exposed to the RF field. Transmission electron microscopy was also able to confirm the internalization of GNPs into these human gastrointestinal cancer cell lines. As seen in Fig. 2, the GNPs in the untreated cells appear to be within endosomes.

Gold salts have been utilized as an immunomodulator for decades in the United States, but they are not considered cytotoxic 26. GNPs are particularly interesting as a therapeutic target for non-invasive RF because a number of gold preparations are already used in clinical practice. Intramuscular gold and oral gold compounds are already approved for use by the Food and Drug Administration as a therapeutic agent for rheumatoid arthritis 2526. These gold formulations used to treat rheumatoid arthritis are well tolerated in a majority of patients 27. Parenteral gold typically causes side effects in about 35% of patients, which can include dermatitis, diarrhea, or stomatitis. More severe reactions such as nephritis, bone marrow suppression, colitis, and hepatotoxicity are more rarely observed 2829. While the toxicity profile for colloidal gold and GNPs does not demonstrate any hematologic or biochemical sequelae, these gold formulations are not currently used in the treatment of rheumatoid arthritis 30. Our findings here are consistent with reports describing the current therapeutic use of gold for rheumatoid arthritis with no apparent cytotoxicity to our cell lines in vitro 263132. However, the potential systemic toxicity of GNPs in humans is not currently known and requires further preclinical investigation before these molecules are deemed safe for clinical trials. We believe this approach is promising and have initiated preclinical toxicity studies of GNPs in normal and tumor-bearing animals.

Once the GNPs are internalized, they serve as target molecules to produce increased intracellular heat when exposed to the external RF field. The PI FACS data displayed in Table 1 and Figure 3 demonstrates the increased percentage of cell death in the GNP-treated cells exposed to the external RF field. TEM reveals disruption and destruction of normal intracellular structures and architecture. Importantly, the difference in RF-induced cytotoxicity between the GNP-treated group and control cells is significant, with over 98% cell death in both Panc-1 and Hep3B GNP-treated groups. The cytotoxicity noted in the control cells is related to non-specific ionic stimulation and heat production that is known to occur in powerful RF fields 33. It will be important to study our system carefully to determine the optimal duration of RF exposure, use of pulsed RF, and RF field strengths necessary to produce lethal injury in GNP-laden malignant cells while avoiding RF-induced damage to normal cells. The current experiments indicate that GNPs are suitable targets for RF-induced thermal destruction of cancer cells. It is possible that shorter duration RF exposures may be sufficient to produce apoptosis-inducing injury in cancer cells bearing GNPs while sparing adjacent normal cells not containing GNPs. To achieve this goal, methods to deliver the GNPs exclusively or preferentially to the cancer cells must be investigated.

It is clear from our data that as an intracellular target molecule, GNPs release substantial heat in the nanoenvironment after exposure to a high-voltage focused RF field. This heating occurs very rapidly (as quickly as one minute) in vitro. The amount of heating related to the intracellular GNPs represents a marked difference compared with the ion rich control samples which contain DMEM and 10% fetal calf serum, but no GNPs. The GNPs in the current experimental system are acting as nonspecific target molecules. Future experimental steps include wrapping the surface of GNPs with a targeting agent to selectively deliver GNPs to malignant cells followed by generation of hyperthermia using non-invasive RF. In this respect, the surface area of GNPs is an important factor for surface functionalization. The reason for selecting GNPs as a target for this study is manifold: 1) recently, GNPs have been used in various biomedical applications 1315192223343536373839404142; 2) as mentioned earlier, colloidal gold and gold compounds have a long history of use in humans 4344; 3) they are easy to synthesize and characterize due to the presence of a characteristic surface plasmon resonance (SPR) band (absent in all other organic based nanoparticles systems such as polymeric nanoparticles, liposomal nanoparticles, dendrimeric nanoparticles) 23; 4) their surface chemistry is relatively simple and surface modification (attaching biomolecules including proteins/antibodies, drugs, and DNA) can be done fairly easily 4546474849 than other relevant technologies (liposomal, polymeric, etc); 5) they have high surface area that allows multiple drug loading on a single particle, and most importantly, 6) they are biocompatible and do not elicit toxic effects 223041505152. Recent in vitro and in vivo reports have confirmed the absence of chronic biochemical and hematological toxicity in mice up to one year after injection of GNPs (1.9 nm in diameter) 30. All of these qualities associated with GNPs make it a potentially ideal molecule for targeted hyperthermia.

We selected GNP of ~5 nm diameter due to the simple synthesis process and high surface area with this size. A spherical GNP of 5 nm size has 23% surface atoms, whereas a 10 nm particle has 11.5%, a 50 nm particle has 2.3% surface atoms and a 1000 nm particle has only 0.2% surface atoms 5354. Due to this higher surface atoms feature, a 5 nm particle will have maximum loading capacity with a minimum gold content. Furthermore, the small size of these nanoparticles may allow them to escape uptake by mononuclear phagocytic cells and penetrate through the smallest capillary pores within the human vasculature.

The preliminary findings here are promising for the use of GNPs as a heat-releasing substrate for this completely non-invasive RF technique. Development of an in vivo tumor model will be important to establish this technique as a feasible treatment modality for solid tumors. Additionally, specific targeting of the GNPs, either through antibodies, peptides, or other entities will likely be necessary to provide tumor-only destruction by the RF and thus, provide significant advantage over current invasive radiofrequency technology.

Our preliminary studies here indicate that GNPs added to the media of human cancer cells in vitro are taken up and localized in vesicles in the cytoplasm of the cells. The presence of these GNP-laden vesicles has no apparent cytotoxic or anti-proliferative effect on the cells. Furthermore, GNPs exposed to an external, non-invasive 13.56 MHz RF field release significant amounts of heat, in fact often sufficient to raise water temperatures to the boiling point. Exposing GNP-bearing human cancer cells to this external RF field in vitro produced dose-dependent lethal injury in > 96% of the cells. Based on these promising results, we have initiated studies to evaluate in vitro cytotoxicities of GNPs and methods to target the GNPs to tumors in vivo to affect RF-induced thermal destruction of malignant tumors.

The author(s) declare that they have no competing interests.

CJG participated in the conception of these studies, performed cell studies and established heating rates of GNPs in solution, analyzed data and wrote the manuscript. CP performed TEM of cancer cells receiving GNPs. RB performed TEM of cancer cells receiving GNPs. PM provided GNPs, participated in the conception of these studies, and participated in writing the manuscript. SAC conceived experimental design and studies, analyzed data, and participated in writing this manuscript.

All authors read and approved the final manuscript