Progress on utilizing hyperthermia for mitigating bacterial infections

Abstract

Recovery from systemic or local bacterial infections can be lengthy and costly, with the clinical challenges being further complicated when bacteria acquire resistance to current antibiotics. Hyperthermia offers new mechanisms for removing bacteria via ablation, or sensitising them to chemical agents. The first part of this review provides a background on the bacterial biofilms, their response to hyperthermia, and acquired resistance to antibiotics, followed by the clinical challenges they present in managing infections associated with soft tissues and biomedical implants. The second part of the review discusses the thermal modalities used to combat infections, including radiofrequency, ultrasound, high-intensity focussed ultrasound, microwave thermotherapy, and photothermal and magnetic nanoparticles (NP). The overall aim of this review is to demonstrate the tremendous potential of hyperthermia for mitigating bacterial infections and foster new research ventures to help remedy these challenging occurrences.

Keywords:

• Bacteria
• soft tissue skin infections
• radiofrequency
• microwave ablation
• ultrasound
• nanoparticles
• non-oncologic applications of hyperthermia

Introduction

Hyperthermic therapies, classically defined as elevated temperatures applied locally, regionally, or to the whole body, for specific time intervals, are routinely utilised in current medical practice. Heat can benefit a wide range of ailments, including arthritis, muscle spasms, weight loss, cancer and healing of cutaneous wounds. Most studies revolve around hyperthermia treatments for oncology; however, recent studies have indicated that similar benefits are possible in the realm of treating local or systemic bacterial infections. Throughout the scope of the present review, both mild (T < 45 °C) and ablative (T > 45 °C) hyperthermia will be discussed for the unique benefits each temperature regime contributes to mediating bacterial infections.

Bacteria, biofilms and their response to hyperthermia

Most pathogenic bacteria are mesophilic and thrive at temperatures between 33 and 41 °C [1]. Elevated temperatures inhibit bacterial proliferation and mobility, which in turn, can increase autolysis and cell wall damage [2]. As demonstrated by Menezes and Teixeira, the outer membrane of Gram-negative Escherichia coli becomes reversibly disrupted above 46 °C, allowing the photodynamic sensitiser, methylene blue, to enter bacterial cells and induce DNA damage, even in the absence of light stimulus, leading to a reduction in surviving bacteria in response to increasing temperature [3]. This study, along with those of Mackey et al., confirms that heat increases the permeability of prokaryotic cells [4–6]. Additionally, exposure of E. coli to 44 or 45 °C for 10 min has been shown to decrease protein synthesis [7].

Biofilm formation occurs when multiple bacteria adhere to a surface and envelop themselves in polysaccharides, proteins, DNA and RNA to produce material that provides a structural extracellular matrix, which protects against antibiotics, immune system responses or other foreign invaders [8–10]. Biofilm integrity is dependent upon concentration and connectivity of matrix components, allowing for development of a highly hydrated structure with high tensile strength that facilitates close proximity of bacteria [11–13]. Compared to planktonic bacteria, biofilms are one hundred times more resistant to antibiotics [14–16]. Architecture and function of the mature biofilm is critical to understanding part of the failure of antibiotics in pathogenic biofilms. As first described by Stewart and Costerton, there are three proposed modalities by which biofilms become resistant to antibiotics: i) inability of the antibiotic to move beyond superficial layers of biofilm due to the formation of microbial aggregates, ii) evolution of antibiotic resistance in some cells and iii) changes in the biofilm environment including decreased pH and oxygen levels in focal areas [17]. In addition, there is a greater frequency of mutations among bacteria protected in biofilms, leading to greater antibiotic efflux or development of acquired resistance [18,19].

Bacteria exhibit not only resistance to antibiotics, but also acquired tolerance to other stressors including oxygen or nutrient levels, radiation and temperature [9]. Similar to chemotherapy agents which target rapidly dividing cells, many antimicrobials are designed to affect the most rapidly proliferating bacteria [20]. However, within biofilms, some cells are pseudo-dormant (persister cells) and thus are not susceptible to antibiotics [20]. Activity of bacteria is coordinated by chemical signals; once the chemical concentration reaches a specific threshold (quorum), the signal for specified activity is imparted to bacteria [21]. Quorum signalling is not necessarily uniform throughout the biofilm due to different densities of micro-colonies that exist at different depths of the film [17]. The ability for biofilms to adapt to environmental stresses, such as fluid shear, is due to the viscoelastic properties of its protein-polysaccharide matrix [22]. Deployment of polysaccharides, iron chelators, DNA degrading enzymes, peptides and quorum sensing molecules have further been evaluated to alter the structural integrity of the biofilm, thus allowing bacteria to become more mobile for enhanced interface with antibiotics [23–26]. The term “structural pulsing,” using pH changes or application of electric field, was coined by Stoodley et al., as possible mechanisms to pump more antibiotics into biofilms [27]. Application of structural pulsing using chemical means alludes to possibilities of altering biofilm structure by physical means, such as heat or osmotic pressures.

Recent studies demonstrated that elevated temperatures alter staphylococcal biofilms to have a lower elastic modulus, and reducing biofilm stiffness might benefit physical techniques used for biofilm removal (debridement) [28,29]. In a study involving hyperthermia alone, Pavlovsky et al., subjected Staphylococcus epidermidis to 1 h of 37, 45 or 60 °C [28]. They noted that there were structural changes (roughening) in the cell walls of bacteria treated at 45 and 60 °C, and a 100-fold reduction in CFUs for bacteria treated at 60 °C compared to those held at 37 °C [28]. Rheological measures demonstrated that physical changes had occurred in biofilms treated at 60 °C, as they had statistically reduced elastic moduli (10 to >3 Pa) and yield stress (23.3 to >3.9 Pa) [28]. Their results support the hypothesis that physical modalities, like hyperthermia, can disrupt the structural integrity of biofilms, which may benefit the effectiveness of antibiotics. Richardson et al. explored the use of hyperthermia up to 60 °C with and without simultaneous delivery of antibiotics in Staphylococcus aureus, Staphylococcus epidermidis and Klebsiella pneumoniae [29]. As shown in Figure 1(A), there was greater than a 50% loss of bacteria viability with temperatures up to 50 °C for 2 h. The benefits of simultaneous application of antibiotics plus hyperthermia were observed for S. epidermidis above 50 °C, whereas heat alone had the same reduction in bacteria above 50 °C as the combination of heat and antibiotics against K. pneumoniae and S. aureus [29]. S. epidermidis was more susceptible to the dual effect of elevated temperature and antibiotic compared to S. aureus. This result represents one of the challenges of applying hyperthermia for augmenting antibiotics, as different bacterial strains have widely varying susceptibility. Similarly, Sturtevant et al. evaluated the susceptibility of S. epidermidis and S. aureus to the combination of the antibiotic vancomycin, and 2 h of 45 °C, resulting in a decrease in viability, as shown in Figure 1(B) [30]. Hajdu et al. carried out similar studies evaluating the response of S. aureus clinical isolates to the antibiotics daptomycin, fosfomycin, vancomycin, tigecycling and cefamandole at 35, 40 and 45 °C [31]. Hyperthermia plus antibiotic was shown to be more effective for reducing biofilm mass and bacterial colonies, with daptomycin having the most benefit. The authors concluded that their results support application of adjuvant hyperthermia in treating infections stemming from medical implants.

Figure 1. (A) The sensitivity of bacteria to a combination of hyperthermia and antibiotics. K. pneumoniae was treated with 0.6 μg/ml of ciprofloxaxin, whereas S. aureus and S. epidermidis were treated with 4 and 1 μg/ml of vancomycin. Reprinted with permission, from Richardson et al. [29]. (B) Viability of S. epidermidis and S. aureus biofilm-residing bacteria treated with increasing doses of vancomycin and simultaneously subjected to 2 h at 45 °C. Adapted with permission, from Sturtevant et al. [30].

Fever is an immune response to bacterial and viral infections with beneficial biological effects, such as increasing leukocyte mobilisation and ingestion of microorganisms [32]. Kluger and Rothenburg showed that fever has the dual-fold potential to reduce pathogenic bacteria (in this case, P. multocida) and systemic iron, which can further impede bacteria’s ability to proliferate [32]. The mechanisms behind why elevated physiological temperature induces hypoferremia are not known; however, numerous groups have demonstrated that the combination of both hyperthermia and hypoferremia reduces the virulence of bacterial infections and facilitates host survival [32–35]. Host responses to infection may include increases in vascular permeability, immune cell mobility and function, inflammatory cytokines (especially tumour necrosis factor (TNF-α), interleukin-6 (IL-6), interleukin 8 (IL-8) and monocyte chemotactic protein (MCP), and interleukin 1β (IL-1β), elevated temperature (either local or systemic) and vasodilation [36,37]. Most importantly, TNF-α, IL-6 and IL-1β are endogenous pyrogens and their release can initiate the onset of elevated temperature [38]. To elucidate the immune response, Seth et al., showed that wounds colonised with both S. aureus and P. aeruginosa had substantially increased expressions of IL-1β and TNF-α compared to wounds colonised by either bacterium individually [39]. In an alternative model, Jiang et al. infected mice with K. pneumoniae to induce bacterial peritonitis. The mice were housed at 35 °C, resulting in an increased core temperature of 39.2 °C, a significant increase in mouse survivability (from 0 to 50%), and reduction in bacterial load in the peritoneal fluid, blood lung, spleen, liver and kidney tissues [40]. The authors also evaluated the impact of 39.5 °C compared to 37 °C on in vitro proliferation of K. pneumoniae and found no difference between the two temperatures [40]. They concluded that hyperthermia in the animals reduced the infection due to the enhanced host immune response mechanisms that occur at febrile temperatures, rather than the impact being due to thermal negation of the bacteria. This study highlights the importance of the host response to hyperthermia in evaluating the impact of heat on inhibiting bacterial infections; hence, in vitro studies may not provide adequate representation of what happens in vivo.

Clinical challenges

Antibiotic resistance

The rise of antibiotic resistant bacteria poses a serious, global threat that may portend science to a “pre-antibiotic” era. Bacterial infections are responsible for over two million severe illnesses in the United States and are responsible for about 23 000 deaths annually [41]. The Center for Disease Control (CDC) estimated that annually, almost 250 000 patients required hospital care due to non-effective antibiotics [42]. Leading drug resistant pathogens include, but are not limited to, S. aureus (methicillin resistant), S. pyogenes (erythromycin resistant), Clostridium difficile (metronidazole resistant) and the emerging pan-resistant Gram-negative Bacilli [43,44]. Nosocomial wound infections are another burden on the healthcare system, increasing hospitalisation of infected patients for an eight additional days [45]. Common reservoirs of biofilms in hospitals are found on medical instrumentation and mechanical ventilation tubing [46]. Richards et al. reported that for hospital inpatients, 95% infected with a urinary tract infection were from catheters, 86% with pneumonia were connected to mechanical ventilators, and 87% of bloodstream infections were from indwelling vascular catheters [47]. The two most common pathogens on intravascular, orthopaedic and reconstruction implants are S. epidermidis and S. aureus, whereas E. coli is most commonly associated with biliary and urological implants [48]. Overall, medical implants generally have low infection rates, from 1–7% depending upon the type of implant, except for bladder catheters (10–30%), fracture fixation device (5–10%), and heart assist devices (25–50%) [48]. Clinicians are faced with multidrug resistant bacterial infections that can sustain for prolonged time periods, and may be easily transmitted by caregivers, contaminated surfaces or instrumentation, sometimes even following thorough cleaning [49–53].

Tissue infections

Approximately 80% of bacterial infections associated with living tissues and medical devices are linked to biofilms and are difficult to eradicate, resulting in acute or chronic infections in cutaneous wounds [54]. As described by Bowler, and also by White and Cutting, even in the absence of visible infection, colonisation of wound tissue with more than 10^5 CFU/g of tissue leads to impaired healing [55,56]. A study of acute versus chronic wounds, obtained from patients with diabetic foot ulcers, pressure ulcers and venous leg ulcers, revealed that 60% of chronic wounds contain biofilm, whereas only 6% of acute wounds do [57]. Furthermore, biofilms of S. aureus in a wound repair model indicates that biofilms are mature after 24 h, leading to an onset of inflammatory mechanisms that inhibit the formation of granulation tissue [58]. Subsequent cutaneous wound healing is slower when bacteria develop antibiotic resistance, as shown in Figure 2, which in turn, delays re-epithelialisation and increases the chance of tissue damage or morbidity [59–61]. Annually, chronic wounds, such as pressure ulcers and leg/foot wounds, affect more than two million people in the United States, with treatments costing between five and nine billion dollars [62]. Wound biopsies of patients with chronic venous leg ulcers having either P. aeruginosa or S. aureus or both species together were evaluated by Fazli et al. [63]. The authors revealed that swab culturing technique may be insufficient for properly analysing bacterial burden in chronic wounds compared to biopsy technique, specifically because they found that P. aeruginosa reside within deeper levels of the wound compare to S. aureus [63].

Figure 2. Photographs of (A) un-infected or (B) infected subcutaneous wounds in a mice, Gram staining of (C) un-infected and (D) infected wounds, (E and F) Fluorescence staining of biofilm components of un-infected and infected wounds, (G) wound re-epithelialization of wounds infected with S. aureus or S. epidermidis compared to un-infected wounds. Adapted with permission, from Schierle et al. [60].

Both antibiotic resistant S. aureus and S. pyogenes are common causes of skin and soft tissue infections (SSTIs) [44,61]. SSTIs are the bacterial infection of the epidermis, dermis and subcutaneous tissue resulting in an inflammatory response. According to the Infectious Diseases Society of America, SSTIs are classified into five major categories: (1) superficial infections such as cellulitis and impetigo, (2) necrotising fasciitis infections, (3) infections caused by animal bites or contact, (4) infections at a surgical site, and (5) immunocompromised infections [61,64]. Treatment against SSTIs includes broad-spectrum antibiotic treatment accompanied by surgical debridement [64,65]. However, the efficacy of empirical therapy is impeded due to the rise of resistant strains [61]. This is especially problematic for necrotising SSTIs, which are more difficult to diagnose early because much of the tissue destruction develops subcutaneously and early signs are similar to other SSTIs [66]. Misdiagnosis and/or delay in treatment correlates with increased mortality due to sepsis [61,67].

One of the major rationales for lower limb amputation in diabetic patients is unresolvable breakdown of soft tissues, leading to exposure of bone, such as those that occur in pressure and diabetic ulcers, or previously irradiated wound beds [68]. The challenge for treating infectious maladies is the difficulty in discriminating between viable and non-viable tissues, especially since the surgical goal is to remove the infection, while retaining the maximum amount of viable tissues. Adequate debridement and antibiotic delivery still leads to about a 30% recurrence rate, mainly due to the rapidity of infectious spread, necessitating additional revision surgeries [69,70]. Current non-surgical techniques for treating tissue infections include hyperbaric oxygen, ultrasound and application of pulsed electromagnetic fields, but mild hyperthermia should be further investigated as an additional adjunctive therapy [68]. Mild hyperthermia has also been demonstrated to facilitate detachment of in vitro biofilms [71,72]. For example, as shown in Figure 3, Nguyen et al. found that slight temperature increases helped to liberate P. aeruginosa from biofilms, while temperature decreases inhibited release of the bacteria [72]. Temperature variations may be of critical importance in dermal wounds. First, skin tends to be cooler than the core of the body (33.5–36.9 °C) [73]. Second, wounds can have elevated or depressed temperature, depending upon the stage of wound healing and presence of infection, which reduces the activity of fibroblasts and neutrophils, resulting in prolonged time for wound closure [74]. For example, McGuiness et al. found that wounds (from trauma or surgical debridement) had an average temperature of 32.6 °C [75]. Further complicating treatment, deep wounds (ulcers, burns, trauma, etc.) may have inadequate vasculature, thus maintaining cooler temperature [76]. It is well established that hyperthermia improves tissue vascularity and oxygenation, and hence might prove to be beneficial in healing infected tissue.

Figure 3. Detachment of P. aeruginosa biofilms grown at the indicated temperature and then increased or decreased to a subsequent temperature, using magnetic heating via iron-oxide nanoparticles, for specific periods of time. Adapted with permission, from Nguyen et al. [72].

Hyperthermic modalities to treat bacterial infections

Radiofrequency ablation

Radiofrequency is a minimally-invasive modality that utilises the electric field created between, or around electrodes to induce localised frictional heating of a target tissue, and is often employed to treat liver, lung, kidney and bone tumours [77,78]. Recently, plasma mediated bipolar radiofrequency (PBRA) (also called Coblation^®) has been used for soft tissue removal, such as for tumour resection and ear, nose and throat procedures [79–81]. Thus PBRA could be rapidly evaluated in clinical trials for enhanced reduction of bacterial burden in removing infected tissue. PBRA generates heat which may contribute to cell death but, in addition, also generates hydroxyl radicals near the applicator which can impart damage to the bacteria [82]. Improvements in wound healing and reduction of bacterial colonisation were evaluated by Nusbaum et al. who compared the effects of wound debridement techniques, including hydrosurgery, plasma-mediated bipolar radiofrequency and sharp debridement on biofilms in deep dermal porcine wounds infected with methicillin-resistant S. aureus, with PBRA providing the greatest reduction in bacterial load [83]. In a similar study, Yang et al. also confirmed that radiofrequency debridement is superior to electric knife (cautery) debridement in not only reducing bacterial burden, but improving re-epithelialisation and collagen synthesis in ablated full thickness S. aureus infected dermal wounds [84]. RFA led to an 85% healing rate, and the authors proposed that the combination of thermal heating and hydroxyl radicals resulted in increased bacterial death [84]. The PBRA technique was also applied by Sönnergren et al. to evaluate the reduction of S. aureus, P. aeruginosa, E. coli and the fungus common in chronic leg ulcers, C. albicans, leading to a 99.87–99.99% reduction in viable micro-organisms amongst the strains [82]. Radiofrequency ablation has also been demonstrated to be beneficial for reducing bacterial growth of S. aureus, Streptococcus and Neisseria associated with infected tonsils [85]. An additional benefit of RFA may be the impact alternating electric current exhibits on the polar molecules of the biofilm components to weaken the biofilm structure [86,87]. The described studies highlight the advantages of reduced bacterial burdens when RFA is used for tissue removal, compared to other non-thermal debridement techniques.

Microwave ablation

As described previously, biofilms have a very high percentage of water, making them susceptible to absorption of microwave wavelengths. Although both radiofrequency and microwave modalities have the potential to be used for reducing diseased tissue, they have different areas over which they are effective, with microwave having a larger area (up to 2 cm), compared to radiofrequency (only a few mm) [88]. Additionally, the potential of radiofrequency becomes reduced when water vapours and tissue charring increase the tissue impedance over during application [88]. While microwave ablation has not yet been used to directly treat bacterial infections in humans, it has been widely used to kill food-borne pathogens, including S. aureus, L. monocytogenes, S. typhimurium and E. coli [89–92].

Although both thermal and non-thermal effects have been shown to be lethal to various bacteria, their mechanisms of action are vastly different. The thermal effects of microwave ablation are a result of increasing the kinetic energy of surrounding water molecules, whereas non-thermal effects can be attributed to the production of an electromagnetic field in the molecules [93]. An alternative mechanism by which microwave irradiation influences bacteria is by changing concentrations of metabolic enzymes such as glucose-6-phosphate dehydrogenase, lactate dehydrogenase, alkaline phosphatase or fibrinolytic enzymes [89]. Sahin et al. was the first done the study to observe the non-thermal effects of microwave radiation (maximum temperature of 44 °C) on bacteria causing prostatitis, specifically E. coli and Enterobacter cloacae, leading to a 50% reduction of colony forming units [94]. Some of the challenges of microwave heating include: (i) uneven heating due to microwave field distributions and the physical and electrical nature of the sample, (ii) inability to control temperatures of microwave-heated samples and (iii) uncontrolled concentration of solute due to evaporative losses from the sample during heating [95].

Ultrasound and high intensity focused ultrasound (HIFU)

Unlike standard ultrasound, High Intensity Focused Ultrasound (HIFU) utilises high amplitude pulses with low duty cycles which minimises heating surrounding structures, while inducing locally high efficiency cavitation to cause coagulative thermal necrosis. It is commonly deployed by physicians due to its predictability and well understood thermal mechanisms for treating cancers of the pancreas, liver, breast, uterus and prostate, as well as soft tissue sarcomas [96–98]. While microwave ablation has both a thermal and electrical effect, HIFU’s mode of action involves a thermal and mechanical effect [98]. There are two modes of cavitation that arise from ultrasound: i) when cavitation bubbles maintain stable oscillation and ii) inertial, in which the oscillations are disorganised, with the bubbles expanding and collapsing [99]. Inertial cavitation further disrupts biological systems via the release of high velocity streams that can damage cells and disrupt biofilm architecture [99]. Utilizing a well-established infection model, Rieck et al. found focussed ultrasound to be beneficial in reducing S. aureus load in dermal abscesses in mice [100]. Either moderate (∼52 °C) or high (∼64 °C) temperatures were employed, leading to a 1.5 or 95% decrease in viable bacteria, compared to the untreated mice. Iqbal et al. explored the application of HIFU for reducing Enterococcus faecalis biofilm and found minimal thermal increases (1 °C over 120 s) but complete eradication of viable bacteria and a corresponding 85% reduction in biofilm thickness [99]. Bigelow et al. investigated the efficacy of HIFU (6.2 and 7.6 MPa) ablation to eradicate E. coli and at the higher exposure, the biofilms were disrupted and had no detectable colony forming units [101]. E. faecalis are not only common in covering urinary catheters, but also in dental infections due to the inability to achieve full disinfection within the uniquely shaped dental spaces [102,103]. To circumvent this challenge, application of ultrasound has been evaluated to break-up oral biofilms and enhance effective irrigation of antibiotics, especially in preparing dental root canals [104,105]. Technological advances have promoted application of HIFU over ultrasound for focal disruption of oral biofilms and delivery of antibacterial nanoparticles (NP) [106].

Although ultrasound may disrupt biofilm, it may not effectively kill indwelling bacteria, hence the promising rationale for coupling ultrasound with simultaneous delivery of antibiotics. There have been a number of studies evaluating ultrasound to facilitate antibiotic delivery in biofilms [107–109]. Ultrasound might impact the effectiveness of antibiotics not only by enhanced transport into the biofilm, but also by promoting oxygenation to enhance bacteria’s metabolism of the agents [110]. An excellent example of the benefits of applying US with antibiotic for reducing E. coli biofilm on a model implant material in a rabbit model was presented by Rediske et al. [108]. US (28.48 kHz, 100 or 300 mW/cm²) was applied continuously with or without systemic delivery of gentamicin, leading to a 2.6 log reduction of bacterial colonies observed with application of the higher energy [108]. Visualisation of E. coli treated with US, as shown in Figure 4, indicates that the cell wall is roughened and pores may form following US treatment [111]. In addition, the fibres extending between E. coli for cell-cell communication appear to be significantly diminished [111]. A notable side observation described by Qian et al. was that even at higher frequency, low-intensity US application did not dissociate the biofilm sufficiently to elicit bacteraemia [112]. In similar fashion, Carmen et al. evaluated the application of ultrasound and vancomycin for treating Gram-positive S. epidermidis biofilms on implanted materials, resulting in a 5 log reduction in CFUs [113]. As described by Vollmer et al. another advantage offered by ultrasound is that since the technique is rapid, bacteria may not have time which to adapt to the applied stress [114]. The promising results from the studies evaluating US and HIFU indicate that there may be many new discoveries in understanding how these techniques can impart damage on bacteria for benefiting clinical application. One of the challenges of translating acoustic modalities for disrupting clinical biofilms is the difference in acoustic pressures at various regions of the transduce (higher near the centre than at the periphery), leading to non-uniformity of the force [115]. Microbubbles might be able to enhance cell killing by further applying cavitation forces to cell membranes (sonopermeability), which would augment antibiotic delivery [116]. Temperature sensitive liposomes have long been explored for localised delivery of high concentrations of chemotherapy. In similar fashion, Wardlow et al. developed temperature sensitive liposomes containing the antibiotic ciprofloxacin to be released at 42 °C, and evaluated the potential of these particles for killing biofilm-residing S. aureus in conjunction with HIFU, leading to a 95% of the antibiotic at 42 °C, resulting in the killing of 90% of the S. aureus [117].

Figure 4. Scanning electron microscopy images of E. coli. Untreated bacteria are shown in part A and bacteria following treatment with 24 kHz, 85 W/cm² ultrasound are shown in part B. Arrow indicates changes in cell wall morphology. Reprinted with permission, from Gera and Doores [111].

Nanoparticles

In recent years, NP have earned an increasingly prominent role in medicine, given their ability to access many parts of the human body due to their small size, and their advantages for biological detection and tagging. Many nanoparticles have been structurally engineered to both target and, upon activation ablate harmful cells or deliver drugs. Here we discuss recent discoveries of silver, gold, carbon and magnetic nanoparticles’ thermal capabilities against bacteria and biofilms.

Photothermal nanoparticles

Photothermal ablation is the process in which a material absorbs light, most often in the near-infrared window (700–1100 nm), and in turn generates enough heat for treatment purposes, specifically, ablative temperatures of >45 °C for eukaryotic and >65 °C for prokaryotic cells. Staying within the near-infrared spectral range is important since the optical absorbers in the body, such as haemoglobin and water, have absorption minima in this range [118].

Silver nanoparticles

Silver nanoparticles are inherently highly effective against microorganisms by inhibiting bacterial growth on medical devices and burns, especially dependent upon their surface area [119–121]. They can be made more anisotropic to facilitate the shift of their plasmon resonance, and hence optical absorption, into the NIR, so that they can be used as photothermal agents [122]. Many NP designed to couple photothermal heating with delivery utilise classic antibiotics. In a slight variation, Hu et al. capitalised on using silver as the antimicrobial agent released from gold nanoparticles following NIR stimulation to kill E. coli. [123]. Most interestingly, only mild hyperthermia (up to 44 °C) was needed to completely eradicate the bacteria [123].

Gold nanoparticles

Similar to silver nanoparticles, gold nanoparticles are excellent photothermal absorbers because their plasmon resonance can be tuned to the near-infrared range [124]. The generation of heat is the major mechanism for ablation of the bacteria; however, as noted by a number of groups, application of intense laser pulses also induces the formation of cavitation bubbles, which may further improve the delivery of antibiotic agents [125–127]. This was demonstrated by Meeker et al. using gold nanocages conjugated to daptomycin for targeted photothermal and antibiotic eradication of S. aureus and S. epidermidis [128]. Similarly, Pissuwan et al. investigated the effects of antibody-conjugated gold nanorods against the pathogenic parasite Toxoplasma gondii using 100 mW of 650 nm laser irradiation, resulting in a death rate of 82.2% [129]. Both continuous wave and pulsed laser irradiation for PTT have been shown to be effective for killing bacteria; however, as noted by Millenbaugh et al. the use of pulsed lasers may be preferential for inducing cavitation [130]. Zharov et al. also investigated the effects on antibody-conjugated gold nanoparticles on bacteria, using anti-protein A antibody conjugated gold nanoparticles for selective photothermal ablation, using nanosecond laser pulses, of S. aureus leading to greater than 90% bacterial death [131].

Carbon nanotubes

Carbon nanotubes are tubular nanostructures consisting of one single wall (SWNT), double walls or multiple walls (MWNT). In medical applications, SWNT, which were first discovered by Iijima in 1991, are more often evaluated than MWNT due to their smaller size (1 nm diameter and tens of nm in length) [132]. For thermal therapy, carbon nanotubes have been shown to be activated by radiofrequency fields, as well as near-infrared radiation [133–135]. Al-Hakami et al. observed the effects of microwave radiation stimulation of carbon nanotubes against E. coli and found that they thermally ablated up to 98 and 100% of bacteria [136]. Levi-Polyachenko et al. demonstrated that photothermal ablation using non-targeted MWNTs and MWNTs targeted with S. pyogenes antibody resulted in a 99.99 and 100% eradication of biofilm residing S. pyogenes [137]. Mocan et al. investigated the use of immunoglobulin G functionalised MWNTs as photothermal agents against methicillin-resistant S. aureus, resulting in an 85.72% reduction of colony forming units [138].

Magnetic nanoparticle heating

Due to their biological biocompatibility, the magnetic iron oxides gamma-Fe₂O₃ (maghemite) and Fe₃O₄ (magnetite), are often explored to act as tissue sensitising agents for inducing hyperthermia (in response to applied alternating magnetic fields (AMF)) [139]. A comprehensive review on the physical mechanisms by which magnetic nanoparticles induce heat is provided by Dennis and Ivkov [139]. Application of AMF causes magnetic moments of the material to be coerced into alignment of the magnetic field. The frequency of the AMF can be rapid enough so that alignment of the magnetic moments lags in time, resulting in hysteresis losses, which generate heat [139]. Frequency of the AMF and field strength are the two variables on which improving the specific loss power (SLP) can depend, but it also depends on the size, shape, structure, surface anisotropy and aggregation of MNPs [140–142]. Independent of field frequency, MNPs with diameters ∼15 nm have been shown to be the most effective for generating hyperthermia [143,144]. For example, Kim et al. utilised an in vivo application of magnetic nanoparticles conjugated to antibodies to increase specificity to S. aureus for localised ablation [145]. The authors utilised a higher frequency and higher amplitude alternating magnetic field than what is typically used for cancer therapy (∼2 MHz>∼500 kHz), and demonstrated that their technique reduced bacterial burden in a wound bed and enhanced healing [145]. Rodrigues et al. used 10 nm Fe₃O₄ nanoparticles to demonstrate the benefits of magnetic hyperthermia over direct heating in the reduction of planktonic and biofilm P. flurorescens [146]. Direct heating to 65 °C resulted in a 3 log reduction of planktonic bacteria and no statistical difference in biofilm residing bacteria compared to the unheated controls. Use of the magnetic nanoparticles to induce 65 °C of the bulk solution in 16.9 min significantly reduced biofilm-residing bacteria beginning at 45 °C and completely eradicated planktonic bacteria at 55 °C [146], as shown in Figure 5. The authors suggest that magnetic heating is superior to direct heating because the nanoparticles impart thermal energy more closely to cell membranes [146]. Similar magnetic hyperthermia technique has been applied to target the pathogenic fungus Candida albicans, which has limited susceptibility to chemical anti-fungals. Chudzik et al. used Fe₃O₄ functionalised with an antibody to target C. albicans [147]. Application of untargeted MNPs for 1 h to induce 43 °C resulted in a 40% reduction of CFUs, and 1 h application to induce 55 °C reduced CFUs by 99.98%. Elevated temperatures of 45 and 55 °C using the antibody targeted MNP did not reduce CFUs as greatly, and the authors propose that the elevated temperatures may be damaging the bond linkages between the NP and targeting agent, causing a delocalisation of the MNPs away from the bacterial surface [147]. Nguyen et al. used iron oxide nanoparticles to produce hyperthermia and disperse biofilms, and in combination with the antibiotic gentamicin, to kill the liberated bacteria [72]. The authors found a statistically significant reduction of CFUs for both planktonic and biofilm-residing P. aeruginosa after 20 min of applied hyperthermia with a T∼9 °C [72]. Although application of T∼17 °C for 20 min resulted in a 69.2% reduction of biofilm mass, this may not be a technique that can translate to clinical application, as the elevated temperature for this length of time would cause severe damage to the host tissues. In addition, the time for most pathogenic bacteria to replicate is 20 min, such that application of hyperthermia over the replication time may allow for development of a thermotolerance response.

Figure 5. Confocal scanning microscopy images of P. fluorescens biofilms treated with direct heating (DH), or heat generated by magnetic nanoparticles (MH). Green color indicates viable bacteria and red indicates dead bacteria. Adapted with permission, from Rodrigues et al. [146].

Yu et al. was the first study to demonstrate the potential of iron oxide nanoparticles to serve as photothermal agents against both Gram-positive and -negative and antibiotic-resistant bacteria [148]. When exposed to near-infrared light, the iron oxide/alumina core/shell magnetic nanoparticles significantly inactivated over 95% of the bacteria after a bulk temperature increase of 18 °C that occurred over 10 min [148]. A unique approach was developed by Galanzha et al. who developed multifunctional gold and magnetic nanoparticles targeted to S. aureus using antibodies directed to protein A or lipoproteins of the bacteria [126]. The authors deployed their technology using a flow system in which circulating bacterial cells could be magnetically collected, enumerated by photoacoustic detection and photothermally ablated [126]. As systems for extracorporeal blood treatment (kidney dialysis, extracorporeal membrane oxygenation (ECMO)) already exist, it is possible that the proposed technique using targeted nanoparticles could be employed to eradicate bloodstream infections.

Conclusion

Both mild and ablative hyperthermia are routinely used in medicine for tumour resection or sensitisation. As demonstrated in this review, and summarized in Table 1, current hyperthermia techniques may be able to parlay into the realm of treating bacterial infections, specifically in managing biofilm-related maladies. Specifically, radiofrequency ablation has enhanced benefits for debridement of infected tissue and hence may be the first modality to become commonly employed. Although ultrasound has been shown to augment the delivery and effectiveness of antibiotics, the prolonged time of application, as well as pressure variances over which the transducer is applied, are limitations. Consequently, HIFU has been utilised instead, and appears to have great potential for both augmenting antibiotics, by localising antibiotic delivery via rupture of carrier liposomes, and for focussed eradication of bacteria in tissues. Usage of microwave thermotherapy for mediating bacteria in tissues seems to be further from widespread clinical translation, although demonstration of its ability to remedy bacterial prostatitis indicates that there may be additional opportunities for exploration. There are also a number of studies highlighting the potential of magnetic and photothermal NP for targeting select bacteria within either blood or solid tissues. Currently, only silver nanoparticles are commonly used in clinical practice, although not yet for generating hyperthermia. Nonetheless, silver, gold and iron oxide NP have limited systemic toxicity in low doses, and as exploration of their utility in cancer extends to treating bacterial infections, they may one day soon be deployed as bactericidal, hyperthermia agents. Which of the currently available techniques becomes commonplace first will most likely depend upon physician preference and a broader range of clinical studies demonstrating efficacy. The field of hyperthermia is on the forefront of understanding new thermal mechanisms, and implementing novel solutions, to combat bacterial infections, especially resistance to antibiotics.

Table 1. Summary of hyperthermic modalities that have been evaluated for eliminating specific microorganisms, including use of the current clinical options: PBRA, RFA, microwave, HIFU, and ultrasound, in addition to hyperthermic nanoparticles which are currently being evaluated in pre-clinical models.

	Microorganism	Reduction	Citation
Hyperthermia modality
PBRA	MRSA in porcine dermal wounds	2–3 log CFU/g	[82]
RFA	S. aureus in dermal wounds	2 log CFU/g	[83]
PBRA	S. aureus, S. pyogenes, P. aeruginosa, E. coli, C. albicans	99.87–99.99%	[81]
RFA	S. aureus, Streptococcus species, Neisseria, all from tonsils	60–100%	[84]
Microwave (up to 44 °C)	E. coli, E. cloacae	50%	[94]
Microwave (63 °C)	S. typhimurium	5 log	[91]
Microwave (74 °C)	L. monocytogenes	100%	[90]
Microwave (73 °C)	E. coli	100%	[92]
HIFU	E. faecalis	100%	[99]
HIFU (non-thermal)	E. coli	100%	[101]
US (non-thermal, with gentamicin)	E. coli	2.6 log	[108]
US (non-thermal, with vancomycin)	S. epidermidis	5 log	[113]
Focused US	S. aureus	95%	[100]
HIFU plus ciprofloxacin- containing temperature  sensitive liposomes	S. aureus	90%	[117]
Hyperthermic nanoparticle
Au NP with release of Ag (temperature increase to 44 °C)	E. coli	100%	[123]
Au nanocages daptomycin (∼55 °C)	S. aureus, S. epidermidis	Below detectable limit (∼100%)	[128]
Au nanorods	T. gondii	82.20%	[129]
Au nanospheres	MSSA, MRSA	69% for MSSA 42% for MRSA	[130]
Au nanospheres	S. aureus	90%	[131]
MWNT	E. coli	98–100%	[136]
MWNT	S. pyogenes	99.99–100%	[137]
MWNT	S. aureus	86%	[138]
Fe3O4	S. aureus	70–80%	[145]
10 nm Fe3O4	P. fluorescens	100%	[146]
Fe3O4 (40 °C or 55 °C)	C. albicans	40% at 43 °C 99.98% at 55 °C	[147]
MNP with gentamicin (▵5 °C)	P. aeruginosa	2 log	[72]
Fe3O4/alumina core shell (▵18 °C) Stimulated with NIR	Gra-positive, Gram-negative, antibiotic resistant variants	95%	[148]
MNP, Au nanorods, Au-coated carbon nanotubes	S. aureus	10–12 fold reduction	[126]

Disclosure statement

No potential conflict of interest was reported by the authors.