Hypoxia, HIF-1

The growth of tumours beyond a critical mass >1-2 mm³ (10⁶ cells) is dependent on adequate blood supply.^4,5

Up to a distance from host vessel of 100-200 μm the initial foci of neoplastic cells receive their nutrients and oxygen by diffusion. Beyond this distance, hypoxia occurs and the need of adequate blood supply is crucial.^5-7 However, the establishment of this neovascular supply in the attempt to overcome hypoxia is inefficient and irregular. It may not occur at the same rate as the proliferation of the tumour. The result is the persistence within the tumour mass of heterogeneous microregions of nonproliferating hypoxic cells, which are surrounded by vital well nourished and proliferating cells (fig. 1).^4,5,7

Figure 1

Photomicrograph of a section of mammary tumor, developed by a 5 month old female transgenic MMTV-neu (erbB2) mouse. T = tumor cord; N = necrosis; Hyp = hypoxic, perinecrotic, rim of tumor cells; C = capillary. Hematoxylin and Eosin. 40x objective.

Several methods to measure hypoxia are currently available and have demonstrated the presence of hypoxic cells both in experimental and in human tumours.⁸

The hypoxic microenvironments are characterized by low oxygen tension, low extracellular pH_e, high interstitial fluid pressure, glucose deficiency, multidrug resistance, increased extracellular lactate concentration and tendency to metastasization.⁴

The forming of new blood vessels depends on a balance between angiogenesis inhibitors and promoters.^6,8 Tumor regional hypoxia and hypoglycaemia are the principal stimulators for the expression of local proangiogenic cytokines, especially vascular endothelial growth factor (VEGF) (fig. 2).^9-11 The early response gene that produces hypoxia inducible factor-1 (HIF-1) and its subunits (HIF-1α and -1β) regulate VEGF expression. HIF-1 is a protein of 120 KDa, member of the basic helix-loop-helix superfamily transcription factors, and its expression is very sensitive to oxygen concentration (1% O2).^11-13

Figure 2

In this figure the structural and functional effects of Hypoxia, HIF-1 and VEGF on tumor microcirculation, cancer metabolism and therapies are illustrated. The vicious circles that occur are also shown. (Modified with permission from: Baronzio et al. (more...)

The adaptation to hypoxia by earlier proliferating neoplastic cells results in the induction of genes that regulate the anaerobic metabolism, nitric oxide synthase and the angiogenesis process.^9,12 Recent studies have shown that HIF-1and VEGF transcripts are overexpressed by several human neoplastic cells including breast, prostate, gastric, colon, lung, bladder and endometrium and they are more active in hypoxic and necrotic areas.^14-17 VEGF is correlated to vascular density, especially in brain tumors and it is associated with bad prognosis.¹⁵ VEGF or vascular permeability factor (VPF), is a 32 to 44-KDa multifunctional potent stimulator of endothelial cells.^13,18 It becomes active by binding to three high affinity tyrosine kinase receptors [VEGFR-1(flt-1), VEGFR-2(KDR/Flk-1) and VEGFR-3(Flt-4)], that are highly expressed on endothelial tumour vessels but not on mature vessels. They exert different effects on endothelial cells (ECs). VEGFR-1 mediates cell motility of ECs, whereas VEGFR-2 regulates vascular permeability and VEGFR-3 lymphoangiogenesis.^13,14,19

Hypoxia appears to be the principal stimulus for the production and the stabilization of VEGF and VEGF mRNA, but recent evidences suggest that VEGF expression has increased in many tumours, as well in absence of hypoxia. This increase results from at least three factors: (A) loss of function of some tumour suppressor genes (such as the von Hippel-Lindau (VHL), p53, p16); (B) activation of oncogenes (including raf, ras,HER2/erb2 (neu) and src^15,20-22) (C) excessive quantity of growth factors, produced by tumour cells and their supporting network (fig. 2).^14,16

Tumor Neovascularization

The acquisition of new in-growing vessels may occur by different mechanisms. ECs are normally quiescent and tightly regulated by a delicate balance between proangiogenic and antiangiogenic molecules.^14,15 In presence of an excessive secretion of angiogenic molecules by tumours, ECs are stimulated and they organize themselves in vessel structure through multistep sequential and distinct processes, depending on tumour type and anatomic localization. These processes include vessel Cooption, vasculogenesis and angiogenesis.

Tumor Vascular Morphology-Perfusion and Hypoxia

Blood vessels associated with the tumor tend to be significantly different in architecture from the surrounding normal tissue (see Table 1)^26-28 and they show, in presence of VEGF and other cytokines, a decreased expression of leukocyte-endothelial cell adhesion molecules (ICAM-1,VCAM-1, E-selectin ) and an enhanced expression of CD 44.^29-31 The decreased expression of adhesion molecules reduce significantly leukocytes and natural killer cells (NKs) recruitment, partially contributing to the phenomenon of immune-evasion,^30,32 whereas the enhanced expression of CD44 may confer a growth advantage on many neoplastic cells.¹⁵

Table 1

Structural and functional abnormalities of tumor vasculature.

Moreover, VEGF induces tumor neovessels to become leakier and to lose a large quantity of fluids (proteins and other circulating macromolecules) towards the interstitium.^33-36 Fluid accumulated inside the tumor interstitium, known as tumor interstitial fluid (TIF), occupies from 30% to 60% of the tumour volume and compared to normal fluids it has a different biochemical nature.^36,37 TIF retention causes an increase in tumor interstitial pressure (TIFP) (fig. 2).^37-39 TIFP increase goes from tumour centre towards periphery, reaching the value of 50 mm Hg, as Jain and coworkers measured in human and experimental tumours on colon, breast, head-neck carcinomas and metastases specimens from lung, liver and lymph nodes.^40-42

Tumor blood flow (TBF) is determined by arteriovenous pressure difference (Δ Pa-v) divided by a factor η for Z, where η expresses the flow resistance or blood viscosity and Z the geometric resistance.³⁴

TBF = Δ Pa-v/ η Z (1)

Normally in tumors, Δ Pa-v is lower than in normal microcirculation. This is caused by the increased number of arteriovenous fistulae present at tumor periphery, that creates a low resistance pathway at tumour surface and diverts blood from entering the tumour mass. Indicator dilution methods, angiography and experimental studies have confirmed this anomalous behaviour.^34,35 Associated to low perfusional tumour supply, Sevick and Jain have found that viscosity parameters Z and η were abnormally high on tissue isolated mammary Adenocarcinoma (R3230AC) and on 22 carcinosarcoma. The geometric resistance Z is a complex function of vascular morphology (i.e., number of blood vessels, branching pattern, diameter, length and volume), and increases according to tumor size, length and vessel tortuosity.⁴² Tumor neovessels are tortuous, heterogeneous, inefficient and devoid of hierarchisation, as confirmed by different Authors using biochemical, corrosion cast and electronic microscopic studies.^27,28

Tumor blood viscosity η has been found elevated and correlated with the shear rate, hematocrit and circulating blood cells deformability.^34,43

Normally capillary diameter ranges from 3.5 to 20 μm, so at some point along their travel in the capillaries, red blood cells (RBCs) as well as white cells (WBCs) have to undergo deformation to pass through. By contrast in the tumor microvasculature, various factors can modify RBCs and WBCs deformability. Particularly the low oxygen partial pressure, the acidic pH, and the increased concentration of fibrinogen tend to make red blood cells and leucocytes less deformable, and more sticky, thus easily trapped intravascularly and blocking TBF.³⁴

A further modifier factor of viscosity is the local increase of the hematocrit. Several authors^34,36,44,45 have demonstrated that this phenomenon is to be ascribed to the fluid loss operating in tumor capillaries with a consequent local hemoconcentration and further worsening of blood flow.

Blood flow measurements in human cancer show heterogeneous values going from highly vascularized organs, such as brain, and poorly vascularized, such as adipose tissue. Perfusion flow at tumour level is higher or lower than in the tissue of origin, depending on the physiological state of the latter.^34,45 It is higher at the tumour periphery than in the central zone and generally primary tumours are better supplied than metastatic lesions. In most studies, perfusion on rodent tumours decreases with tumour size when compared to normal tissue. However, in the minority of experiments the decreased flow was not confirmed even in a similar tumor type. Several pathophysiological mechanisms have been proposed to explain this difference such as transplantation site, stage of tumor growth, flow registration and recording methods.⁵

TBF is not regulated according to metabolic demand as in the case of normal tissues.⁴ This decreased metabolic adaptability to cells associated to an irregular blood availability (perfusion) produce a clusters of cells, lacking nutrients and oxygen (hypoxic cells).⁴⁶

Two kinds of hypoxic or clusters of cells in situation of low energy state have been recognized: (A) Diffusion-limited or chronic hypoxic cells; (B) Transient or acute hypoxic cells. The two types of hypoxia have different origin and coexist together in a well-perfused zone of the tumour mass too, causing a functional disturbance of macro and microflow.^4,5 A more realistic vision shows that these two situations change continuously, because tumour blood flow is time fluctuating.^45-47

Chronic hypoxia is the result of availability of nutrients and oxygen towards the tumoral tissue, the diffusion in the extracellular space and the respiration rate of cancer cells.^4,5 It has been calculated and observed that when cancer cells are >100-200 μm away from functional blood supply become hypoxic and suffering (fig. 1, fig. 3).

Figure 3

Photomicrograph of a section of mammary tumor, developed by a 5 month old female transgenic MMTV-neu (erbB2) mouse. T = tumor cord; N = necrosis; L = lymphatic vessel; C = capillary. Hematoxylin and Eosin. 20x objective.

Cells adjacent to capillaries displayed a mean oxygen concentration of 2%, located at 200 μm displayed a mean oxygen concentration of 0.2%.⁴⁷ The distance from the nutritive vessels, the haemoglobin concentration and the blood flow crossing that tumour area are the only parameters responsible for chronic hypoxia. In fact, the removal of O₂ by tumour cells, has been calculated to be efficient or better than the one of normal tissues, revealing that neoplastic cells in vivo do not have impaired ability to utilize oxygen as proposed in the past.⁴⁵

Acute hypoxia is the result of intermittent opening and reopening of tumor blood vessels. Among the various factors responsible for this temporarily blood flow stop, two of them seem the most plausible:

1. TIFP combined with the irregular expansion of tumour mass, whose three dimensional growth is subjected to a continuous remodelling in a confined space and it causes a temporary compression or occlusion of some tumour capillaries.^5,37-39
2. transient stop of tumor blood flow or supply by platelets plug (see fig. 4).^5,47

Figure 4

Photomicrograph of a platelet thrombus in a peripheral blood smear of a mouse bearing an Ehrlich carcinoma implanted on the hind leg. MGG staining. X 63 objective. Differential interference contrast.

In our opinion, this intravascular thrombosis deserves to be taken into much higher account than usually done.⁴⁷ In fact, the majority of cancer patients have coagulation abnormalities associated to hypoxia.⁴⁸ Recently, it has been demonstrated that hypoxia not only induces VEGF but also stimulates endothelial cells to over express tissue factor (TF) and Plasminogen activator inhibitor (PAI-I). These factors induce endothelium to become prothrombotic and cause fibrin formation and platelet activation.⁴⁹ Furthermore, VEGF binds to fibrinogen and fibrin by stimulating endothelial cell proliferation.⁵⁰ Fibrin has been demonstrated to be essential for supporting endothelial cells spreading and migration.⁶ The haemostatic system, in a certain sense, becomes a regulator of angiogenesis and it can partially explain acute hypoxia and its regional appearance and disappearance.^5,47,48,51 Concluding hypoxia becomes a self-perpetuating mechanism able to trigger angiogenesis, intratumoral fluid accumulation and thrombosis (fig. 2, fig. 4).⁴⁸

Effects of Hyperthermia on Metabolism

Recent studies have emphasized that metabolic alterations often follow hyperthermia.⁶³ Mueller-Klieser et al,⁶⁴ using bioluminescence and photon counting methods for imaging in situ the metabolites used by tumour tissues, have shown an enhanced glycolytic activity followed by an increase in lactate levels upon heating application.⁶⁵ The higher glucose and lactate concentrations may be the result of a temporarily increase in blood flow and an expansion of interstitial tumour compartment.⁶⁶ Furthermore, the intensified acidosis with the enlargement of hypoxic tumor areas that follows the transient decrease in tumor perfusion application sensitize and facilitate tumor cell destruction by hyperthermia (fig. 6).^67,68

Figure 6

In this figure the effects of HT on hypoxic cells, on tumor microcirculation and on surrounding normal tissue are illustrated. In bold black are shown the direct effects of HT, whereas the indirect effects are shown in gray. (Modified with permission (more...)

Effects of pH on Hyperthermia

Since Gray's studies in the fifties (1955),⁶⁹ it has been demonstrated that hypoxic and acidic regions in solid human tumors are common. These cells, chronically exposed to low extracellular pH, are relatively resistant to ionizing radiation but they tend to be markedly sensitive to the thermal damage.^68,70 Earlier studies, in vitro conditions, ascribed this effect to the low pH of tumor milieu (pHe) and showed that thermosensitization was more evident at mild temperatures (e.g., <42°C) rather than at higher temperatures. Gerweck and coll have demonstrated that environmental parameters such as hypoxia and acidosis, can strongly modify the cellular hyperthermic response. The Chinese Hamster Ovary (CHO) cancer cells survival curves in vitro have clearly evidenced that the oxygen enhancement ratio (OER) for hyperthermic cell kill is ≈ 1 in contrast with 2.5 to 3.0 for radiation inactivation (fig. 7).⁷⁰ From these studies the authors concluded that hypoxic cells appeared to be slightly more sensitive to hyperthermia than oxygenated cells. As hypoxic cells live in acidic environments, these authors studied the same CHO cells in vitro at different pH of medium (see fig. 7). They concluded, in accordance with other authors, that exposure of tumor cells to hyperthermia at conditions of pH below 7.4 enhanced cell lethality. Such increased cell sensitivity, often larger by a factor of 10² at some dose level, continued even during the development of thermotolerance.^69,70 For clinical purposes, Gerweck and Richards⁷¹ have introduced the concept of “pH enhancement ratio” that is the ratio of the inactivation rates at pHe 6.7 vs. pH 7.4. This pH range has been calculated in vitro on different tumor cell lines and according to these authors, it is the only value among which pH sensitisation takes place.⁷¹ These authors refer that the pH sensitising effect is manifest over a pH range which is observed in tumor tissue, i.e., pH 6.6 to 7.0 (fig. 8).^68,70,71

Figure 7

OER curves of Chinese hamster ovary cells (CHO) treated with hyperthermia. Hypoxia was induced 10-20 min prior to treatment. (Suit an Gerweck , Cancer Res 1979; 39:2290-2298. Reprinted form ref. with permission).

Figure 8

CHO cells cultured at different pH under aerobic conditions and heated during the midportion of pH exposure conditions. (Suit an Gerweck , Cancer Res 1979; 39:2290-2298. Reprinted from ref. with permission).

However, heat sensitivity of cancer cells seems to be limited when they are exposed to a chronic low pH_e.⁷² In fact, Hahn and Shiu,⁷³ after studying CHO cells in vitro, maintained in acidic medium for different time before the exposition to heat treatment, have demonstrated that cells exposed to a low pH for prolonged periods were less sensitive to heat than cells exposed to an acidic medium shortly before heating. Other authors have confirmed these observations and have shown that cancer cells exposed for long period to low pH milieu, are less responsible to heat treatment and have a higher pHi compared to a brief period adapted one.^69,70,74 Other experimental studies have demonstrated that it is sufficient an acute reduction of pHe below 7.0-7.2 to increase the hyperthermia damage and to decrease thermotolerance.^69,70,75,76 These studies in vitro have been performed under conditions where pHe was rapidly altered, a condition that does not correspond to one in vivo conditions where low pHe is gradually achieved and the cells are exposed to an acidic environment for more prolonged periods, the so-called chronic hypoxic cells. These phenomena can explain the less dramatic effects on hypoxic cells obtained on patients after a treatment of hyperthermia as reported by van der Berg.⁷⁷ Recent studies have tried to differentiate the pHe sensitising effect in vitro and in vivo. Rhee et al⁷⁸ studied the effectiveness of low extracellular pH in sensitising cells to heat using SCK mammary carcinoma cells of A/J mice. They have demonstrated a different thermosensitivity and thermotolerance following hyperthermia, at different pH values (pH 7.2, 6.6, 5.5) on in vivo and in vitro derived cells. For any heating temperature tested the sensitising effects of pH was much smaller on in vivo derived cells that on in vitro derived cells. Furthermore, this reduction of pH effect was observed for cells derived from larger tumors as well as tumors at an early stage of growth in which the internal milieu was not acidic. This indicates that cellular adaptation to low intratumor pH was not the sensitising factor and might be related to other factors than pH, such as nutrients deprivation and decreased blood perfusion.^78,79

Although recent experiments proved that thermosensitivity is more dependent upon pHi rather than pHe^75,80 pH_e value alone has demonstrated to be a useful prognostic indicator and that the reduction in pH_e induces a decrease in the intracellular pH.^81-85

Since that extracellular pH_e is the result of lactic acid accumulation, von Ardenne⁸⁶ tried to further lower the tumor milieu pH through the administration of supraphysiological levels of glucose. Earlier animal and human experiments indeed demonstrated that intraperitoneal or endovenous injection of glucose reduced intratumor pHe and increased the response to thermotherapy. ^86-88 However, recent studies have evidenced that the change of pH is induced by tumor blood flow reduction or nutrition deprivation rather than by the increased glycolysis (fig. 9).^79,83

Figure 9

In this diagram the direct effects of glucose administration (bold white) on Tumor Blood Flow (TBF) and on tumor extracellular/intracellular pH are illustrated. Associated are shown the indirect effect of hyperglycaemia on cardiac output (bold gray) with (more...)

Although the experimental evidence has demonstrated that hyperglycaemia is a useful method for enhancing the response to thermoradiotherapy, its clinical application is not yet of routine. This is partly due to a fear of inducing uncontrollable physiologic alterations and to an absence of a standardized protocol.^{89- 91} The clinical application and administration of hyperglycaemia will be discussed later.

Introduction

Hyperthermia has as principal goal that of destroying tumor tissue (vasculature included) by achieving a temperature that exceeds the cytotoxic threshold (42.5°C) and induces cell death in tumor tissue with a selective sparing of normal surrounding tissue.⁹²

Although cancer cells are destroyed by hyperthermia alone, many factors, including the cell type and blood perfusion, influence its success. In fact, it has been shown experimentally and theoretically that heat transferred away from a tissue is the result of the rate and the volume of blood flowing through that tissue (perfusion).^{93- 95} Since the importance of blood perfusion in heat dissipation, a brief description of heat effects on tumor blood flow [TBF] and on endothelium is useful.

Effects on Tumor Blood Flow (TBF) and on Normal Circulation

The relationship between temperature rise and perfusion has been demonstrated by Jain et al⁹² to be inversely correlated: as the perfusion rate decreases the tumor temperature increases. The gap in temperature obtainable between normal and tumor tissue is due to the differences in conduction and convection characteristics between the two tissues.⁹⁶

Tumor blood flow and distribution are different from normal tissue and show regional variations in the same tumor itself as quantified by Gullino and Grantham.⁹³ Blood flow in human cancer showed a greater variability as compared to animals, however a different flow between normal and tumor tissue exists.⁴⁶ Tumor blood flow appears to be inferior than that of normal tissue and to have a decreased adaptability to metabolic demands and physical stress (heat).⁴⁵ The reasons are to be ascribed to the absence of innervation.^45,46 The vasodilatation that happens after heat application to normal tissues is not present at the same extent in tumor vasculature.^96,97,98 This determines a decreased convection and permits to entrap heat in the tumor area, rising the temperature in that target area.^97,98 In fact, different authors heating deep seated human tumors by radiofrequency (13.56 MHz) therapy, have reported that tumor temperature was higher than that of surrounding normal tissue.^99-103

After heat exposition macroscopic blood flow measurements in normal tissue such as skin or muscle showed a rapid and a dynamic vasodilatation with increased permeability of the vascular wall (fig. 6). Song^94,95 and Vaupel⁹⁷ demonstrated that the degree of alteration was temperature and treatment duration dependent. The heating changes in the tumor were slightly absent or increased at the beginning of treatment, if the treatment time was prolonged a decrease with a stasis of blood flow took place.^97,98 Similar conclusions on blood flow behaviour during hyperthermia have been reached by other authors using microscopic measurement such as RBC velocity, laser doppler flowmetry and hydrogen clearance methods.^97,98 Jain et al⁶⁸ studied RBC velocity and vessel lumen diameter in mature granulation tissue and in neoplastic tissue (VX2 carcinoma). They found and confirmed the above-mentioned observations, but noted that stasis occurred in normal and tumor tissue both. The difference in stasis was dependent on temperature. In fact, the stasis in normal tissue occurred later and at higher temperature (47°C) whereas stasis in tumor tissue was reached before and at a lower temperature (41.00°C).⁹⁸ As for stasis, Li⁹⁶ founded that the recovery kinetics of tumor blood flow after heating was temperature dependent, i.e., blood stasis occurred in the range 3-5 hr after heating, remaining low for 24 hr and partially recovering after 48 hr. Aside the vasodilatation a complexity of other events follow heat application. They are different biochemical and microcirculatory changes, such as: acidosis, RBC stiffening and aggregation, degenerative changes of endothelium, increased vascular permeability, platelet aggregation, leukocyte sticking and intravascular clotting (fig. 6).^97,98 These phenomena worsen the tumor microenvironment further and explain why neoplastic cells are damaged more easily by temperature (42-45°C) than normal cells.⁹⁵

Effects on Endothelium

Different in vivo and in vitro studies have shown that endothelial cells [ECs] and in particular those proliferating ones can be damaged by heat.¹⁰⁴ The damages can regard the integrity of endothelium or its vulnerability to heat. Histological methods have revealed that after hyperthermia, a rapid reduction and rearrangement of F-actin stress fibers follows. These fibers maintain the junctional integrity among the cells. Their lack determines an increase in vascular permeability, a phenomenon usually observed after HT.^103,104 The effects of hyperthermia are not however similar in all tumor types. This difference has been demonstrated by Nishimura¹⁰⁵ and it is dependent on the quantity of connective tissues present in the vascular architecture. Furthermore Hyperthermia regulates positively different adhesion molecules on endothelium surfaces providing an increased recruitment of T cells to tumors.¹⁰⁶ This effect, associated to the increased vascular permeability and extravasation, partially explains the escape of leukocytes from the vasculature and the peritumoral inflammatory reaction.

As described, tumor blood vessels are more vulnerable to heat than normal surrounding blood vessels, probably for their structurally immaturity.^4,5,95 Furthermore, tumor vessels, as neoplastic cells have shown to acquire thermal adaptation (Thermotolerance) to reheating.⁹⁴ This phenomenon, referred by Song as vascular Thermotolerance (VT), differs as regards neoplastic cells Thermotolerance (NT). NT refers generally to a resistance to heat induced cell killing through the production of Heat shock proteins (HSP), while VT is a thermal adaptation developed by tumor vessels that initially respond to the stress of reheating by increasing blood flow instead of reducing it. The reasons for this behaviour are not completely known and remain actually only speculative.⁹⁴

Antiangiogenetic Effect of Hyperthermia

Hyperthermia has demonstrated to kill tumor cells by a direct and an indirect mechanisms. ⁹⁷ The indirect killing mechanism has been recognized as an inhibitory effect on angiogenesis. In the 60s the inhibition of angiogenesis was ascribed to ischaemia with a consequent obstruction and destruction of tumor blood vessels followed by an inability to form new vessels.^97,104 Recently a biochemical mechanism has been recognized.¹⁰⁷ These Authors have clearly demonstrated that heat application can inhibit angiogenesis by activation of a Plasminogen Activator Inhibitor I-dependent mechanism (PAI-I). The effect has been investigated and found operating in vivo and in vitro with a reduction in the number of microvessels. Endothelial cells viability was not affected.¹⁰⁷

Introduction

Tumor vascular supply is extremely important for maintaining tumor growth, progression and metastasization. Furthermore, tumor perfusion can affect related micro environmental parameters, such as oxygenation status, pH distribution, bioenergetic status, nutrient supply and sensitivity of cancer cells to anticancer treatments, hyperthermia included.^95,96 Blood flow is the major determinant of heat dissipation and it is responsible for the selective and uniform heating increase in tumor target area respect the normal counterpart.⁹² The relationship between blood flow and the effective tumor conductivity has been studied by Jain et al,⁹² who found that temperature rise in tumor mass is inversely and linked to the effective thermal conductivity. Since hypoxic cells and preferentially cells exposed, to an acute acidification have shown an increased thermosensitivity (fig. 7), many investigators have developed different methodologies with the aim of increasing thermoresponse as the heat induced cell damage.⁹⁵

Clinically, a modification of tumor thermoresponse could be obtained or modulating tumor blood flow, tumor microenvironment or trying to ameliorate the heat deposition into the tumour tissue (Table 3, and fig. 10).^{95,97,103,108} Otherwise expressed the clinical attempt to increase the heat response can be reached or by inducing hypoxia through blood flow manipulation or trying to render tumor cells more vulnerable to heat.

Table 3

Clinical methods for improving thermoresponse.

Figure 10

In this figure the interactions of Hyperthermia with the chemotherapy and radiotherapy are illustrated together the principal points of tumor microenvironment modification [TBF (Tumor Blood Flow); Micromilieu, Membranes, Metabolism] for improving the (more...)

Blood Flow Reduction or Deprivation

Cutting off the tumor's blood supply by physical clamping produces 100% radiobiological hypoxia¹⁰⁹ that increases cancer cells death by apoptosis. Field⁸³ has demonstrated drastic changes in the isoeffect relationship when deprivation of blood supply by clamping was applied upon hyperthermia. Firstly, tumor tissue becomes more sensitive, equivalent to a factor near two in heating time; secondly the transition at 42.5°C is eliminated, showing a reduction or an abolishment of thermotolerance.

Drugs Able to Modify Tumor Blood Flow^{95,96,103,108,113}

The goal of TBF modulation by drugs is to make the tumor sufficiently hypoxic/or underfed and consequently more thermo sensible, similarly to clamping or Chemoembolization. Drugs, which are of course less cumbersome of the preceding methods of treatment, are of two types: vasodilatators and agents that attack vascular endothelium (VTA). Vasodilatators starves the tumor through a steal phenomenon, while VTA get tumor hypoxic by disrupting and decreasing the nutritive vessels.

1. Vasodilatators: Hydralazine, Calcium blocking agents (verapamil, flunarizine), Serotonin and its Analogues.^102,103,108
2. Vascular Targeting Agents (VTA): FAA, DMXXA, CA4DP, TNF; IL-1.^{76,103,114,115}

Hyperglycaemia and Drugs Acting on Tumor Metabolism

Hyperglycaemia and drugs act on tumor metabolism.^76,75,,^86,88,92 Several animal studies have demonstrated that hyperglycaemia can sensitise mammalian cells to heat.^116,117,118 Initially the effect was thought to be metabolic induced (pH drop, due to excessive lactic acid production), but recent studies by Ward-Hartley and Jain have demonstrated that the hypoxia and pH reductions are secondary to the tumor blood flow reduction.¹⁰³ Calderwood and Dickson reported that intraperitoneal injection of glucose reduced tumor blood flow by more than 90% for a few hours.¹¹⁶ Furthermore, these authors have demonstrated that blood flow inhibition and pH reduction are related to the serum concentration of glucose level.¹¹⁷ Nagata¹¹⁸ reported similar effects on 25 cancer patients who received intravenous injection of 500 ml of 10% glucose. Tanaka¹⁰² has clearly demonstrated that patients treated with hyperglycaemia were more responsive to thermoradiotherapy than a control group treated by thermoradiotherapy alone. The mechanisms responsible for blood flow reduction in tumors during hyperglycaemia are multiple and have been explained by Ward and Jain.⁹² The reduction is a consequence of systemic and local effects both. The systemic effects were ascribed to a significant cardiac output redistribution that consequently reduced tumour blood flow by steal phenomena (see fig. 9). The local mechanism that contributed to reduce tumor blood flow was linked to Red blood cell (RBC) deformability decrease. Initially this increase in rigidity was postulated to be pH dependent,⁹⁵ but Crandall et al¹¹⁹ have demonstrated that it was necessary a long exposure time to a low pH for obtaining a RBC rigidity. Traykov and Jain questioned this lag period and demonstrated that RBC deformability followed almost immediately the infusion of glucose and galactose.¹²⁰ This rapid local phenomena associated to the steal effect of cardiac output are responsible for the initial blood flow reduction. The metabolic interaction with pH decrease can perpetuate the reduction of blood flow for many hours after glucose administration. Studies by Hasegawa¹²¹ have shown that a difference in pH drop exists between tumor and normal tissue. In fact, 30-60 min after glucose administration the pH rapidly dropped of 0.3 to 0.6 units in tumor, as compared to a decrease by only 0.1 unit of normal tissue. The complete recovery to baseline was also different in the two tissues. Animal studies have demonstrated that the administration of glucose prior to hyperthermia can modify tumor regrowth and thermotolerance (i.e., fig. 11).¹²² Tumor regrowth delay was greater in the group treated with glucose as compared to the group not treated. Thermotolerance disappeared 12 h after heat treatment in the group with glucose administration, whereas in the untreated group thermotolerance disappeared only after 72 h.¹²² The majority of experimental studies have been performed in vitro and in animals. Recently, after the increased use of hyperthermia in association with radiotherapy, many human studies have been performed. Dickson Calderwood demonstrated that serum glucose levels in the range of 1000 mg/dl can cause a complete cessation of tumor blood flow.¹¹⁶ Similar levels of glucose in humans are however not attainable without side effects. Levels of 400mg/dl have been demonstrated by Lippmann et al¹²³ the maximum attainable for long periods (24h) with no modifications of blood count and acid base equilibrium. Krag et al⁹¹ with the goal of achieving a steady level of 400 mg/dl for short periods of time (3h), intravenously administered a glucose loading dose of 77 mg/m² to three patients with advanced metastatic cancer. After 20 minutes, levels over 400 mg/dl were obtained and maintained for 3 h with no apparent side effects and rebound hypoglycaemia. Levels beyond 700 mg/dl were attempted, but could not be maintained without side effects. Other authors have studied and compared the effects on tumor blood flow, pHe and the clinical response of human patients submitted to a glucose load.^124-126 Nagata et al¹²⁵ demonstrated that a glucose administration of 500 ml of 10% glucose by intravenous route reduced the tumour blood flow, measured by laser Doppler flowmetry, to 66% of the baseline level at 30 min after the beginning of infusion. A complete tumor response (CR) of 30% was obtained on glucose treated patients compared to a group treated only by Radiotherapy and hyperthermia.¹²⁵ Leeper et al¹²⁶ have determined whether intravenous (i.v.) or combined intravenous plus oral glucose administration were more effective in inducing acute tumor extracellular acidification. They concluded that the effects of hyperglycaemia induced by i.v. + oral administration were similar and i.v. exhibited an acidification of 0.14 ± 0.002 pH unit after 91 ± 7 min of infusion. Engin and coll.⁸⁵ have evaluated the importance of extracellular pH as prognostic indicator of tumor response to thermoradiotherapy. The authors measured the tumor pH_e of 26 human tumors with a needle microelectrode of 2.5 cm of length. They reported that the difference in pH_e exhibited by complete responding (CR) patients and noncompletely responding (NCR) 6.88 ± 0.09 versus 7.24 ± 0.09 was statistically significant (p≥0.08). On these grounds, they suggest that extracellular pH_e measurement may be a useful prognostic indicator of tumor response to thermotherapy. Although, a pH_e reduction of ≈0.2 units is easily obtainable by glucose load,¹²⁷ greater reduction ≥0.5 units are necessary for inducing acute intracellular acidosis.¹²⁸ The degree of reduction of pH_i that accompanies acute extracellular acidification is the critical factor for sensitizing cells to hyperthermia and for abrogating the heat shock proteins induction. Recently various Authors have demonstrated that for obtaining such pHe drop in melanoma, metabolic inhibitors such as meta-Iodio-benzylguanidine (MIBG) or alpha-cyano-4-hydroxy-cinnamic acid (CNCn) must be added.¹²⁹ For understanding, the biochemical points of action of these inhibitors (see fig. 9). In conclusion, the concomitant administration of glucose together with MIBG increases the tumour magnitude and duration of acidification and the oxygen tension.^130,131 This association has the potential to improve response to radiation therapy and to hyperthermia itself.

Figure 11

Growth curves of BT₄ An tumors after ACNU 20 or 10 mg/kg combined with hyperthermia with or without hyupertonic glucose 6 g/Kg i.p. 2 hours before treatment. It is interesting to note the decreased tumor regrowth after glucose-chemotherapy-HT administration. (more...)

The protocol used in our laboratory was similar to that used by Nagata (Table 4).¹²⁵ Following these studies, we used 500 cc of Glucose at 10% obtaining blood glucose value of 300-400 mg/dl without side effects (unpublished observations). Recently, following the suggestions of Leeper group^129-132 we have added to the hyperglycemia two metabolic inhibitors such as quercetin and Amiloride (Moduretic®) (Table 4).

Table 4

Treatment schedule clinically used by our group.

Modifiers of Thermal Sensitivity

Lidocaine and anesthetics,^133,134 Calcium antagonists,135 Polyunsaturated fatty Acids.¹³⁶ Cycloxygenase inhibitors,¹³⁷ Betulinic acid,¹³⁸ aldehydes,¹³⁹ Vitamins and bioflavonoids.^{139- 142}

Rapid Heating

Rapid heating is a method developed by Hasegawa group¹⁴⁵ in the attempt to shorten hyperthermic treatment still reaching temperature sufficient to kill tumour cells and to change tumor blood perfusion. The experiments have been made on C3H mice inoculated with SCC-VII tumor in the thigh, heated with warm water bath and RF heating devices. C3H mice were divided in two treatment groups and compared, the former in which the heating temperature was increased to the target temperature in 1 min, and the latter group in which the heating temperature was gradually increased. The following parameters were studied: changes in blood flow in tumour and normal tissue, tumour growth rate, cancer cells apoptosis. Changes in blood flow were not observed in the slow heating group before or after the hyperthermic treatment, whereas in the rapid heating group a significant increase in blood flow was observed in the normal tissue followed by a significant decrease after heat treatment in the tumour tissue. Tumor growth delay was more evident in the RF rapid heating group compared with warm water heating group. Apoptosis and cytokinetic activity modifications were favorable to the rapid heating group, revealing that a vascular injury was effectively obtained with a shortage in treatment time in this group.¹⁴⁵ Clinical studies on this methodology are warranted.

Mild Hyperthermia and Oxygenation

As previous described, solid tumors contain regions of low extracellular pH and oxygen that may affect treatment outcome. Laboratory and clinical data confirm that hyperthermia may enhance the therapeutic index of ionizing radiation.¹⁴³ Several mechanisms have been found and are summarized in the recent reviews of Kampinga¹⁴⁴ and Vujaskovic.¹⁴⁶ Among these mechanisms, tumor oxygenation improvement after mild hyperthermia (HT with temperature between 39-42.5°C) is now considered to be of the utmost importance. As determined by Song et al,⁹⁵ normal and tumor tissue show a different behavior following heat deposition. Blood tumor vessels respond markedly different to a second heat application showing a greater vulnerability and vasodilatation to heat than normal surrounding blood vessels. This phenomena, referred by Song as vascular Thermotolerance (VT), appears to account for the improvement in the tumour blood flow observed after the reheating at 42.5°C. As the blood flow increase, an improvement in tumor oxygenation follows which may last for as long as 24-48 h.^94,95 Tumor oxygenation by mild HT has been found to be more effective than carbogen breathing in increasing the radiation response of experimental tumors.¹⁴⁷ Clinical studies on 18 patients with locally advanced breast cancer treated with thermo-chemo-radiotherapy have confirmed these experimental results. Tumour oxygenation improvement appeared to be temperature dependent and associated with the lower thermal doses.¹⁴⁸