Current status of antivascular therapy and targeted treatment in the clinic

Abstract

Antivascular and targeted therapy are now an integrated part of the treatment of myelogenous leukemias, GIST tumours, B-cell lymphomas and breast cancer. In various malignancies improved responses and prolongation of survival for several months is regularly reported. The progress in this field is relevant for hyperthermia. Heat has among other effects documented antivascular effects, and can be considered as one of the established methods in the field based on several randomised phase III studies. Hyperthermia should be considered for combination with other antiangiogenic agents.

• Clinical studies antivascular drugs
• hyperthermia

Introduction

A vast number of reports on the preclinical and clinical effects of therapy directed against angiogenesis (i.e. formation of new vessels) and the vascular network in tumours have been published in the last five years. Currently there are more than 31 000 PubMed hits on the term ‘angiogenesis’, and more than 1 000 hits on ‘bevacizumab’. Awareness of this development is crucial for the judgement of the true value of progress in clinical hyperthermia. This review will highlight some aspects with special reference to randomised clinical trials. Judah Folkman has described the crucial role of the vasculature for tumour nutrition and growth. He initially postulated the ‘tumour angiogenic factor’, a concept which has been developed into a cascade of factors which induce or inhibit the formation of new vessels [1–4]. Angiogenesis is now acknowledged as one of the ‘hallmarks of cancer’, and proper function of blood vessels is a prerequisite for delivery of drugs to solid tumours [5], [6].

Targeted therapy

The term ‘targeted therapy’ was originally introduced when monoclonal antibodies customised against specific cellular targets became available in the 1980s [7]. Today, the term includes all molecules that target different signal transduction elements, or where the mechanism of action is thought to be at a particular molecule [8]. Thus, it includes all antiangiogenic and antivascular therapies where the specific target molecules are known. It is interesting to note that the term by itself is a seducing one, as it implies that we actually know the mechanisms which are blocked or stimulated. Research in this field was boosted by the overwhelming success of imatinib which rapidly became accepted as standard therapy for chronic myelogenous leukemia worldwide [9], [10]. The discovery that the Philadelphia chromosome in chronic myelogenous leukemia consisted of a translocation between chromosome 9 and 22, leading to a constitutive activation of the Abelson tyrosine kinase (Figure 1), which could be inhibited in culture and also in mice, paved the way for clinical testing [11–17]. In addition, imatinib also inhibited the oncogen c-kit, and therefore proved to be especially effective in gastrointestinal stromal tumours (GIST) [18]. The effect in these tumours was better than historical controls treated with doxorubicin [19]. Imatinib also exhibited an inhibitory effect on the platelet derived growth factor receptor (PDGFR) in vitro, and this proved to be effective in dermatofibrosarcoma protuberans where a translocation of chromosome 17 and 22 leads to autocrine activation of platelet derived growth factor β-chain (PDGF β-chain) [20], [21]. In this neoplasm, imatinib induces complete or partial response in most patients treated for advanced disease. In eosinophilic leukemia and hypereosinophilic syndrome, different translocations lead to fusion products that involve PDGFR-α or PDGFR- β, both responding well to imatinib [22–27]. An important lesson from the use of imatinib is that resistance can develop by mutations in the target molecules [28]. Interestingly, the new drug dasatinib is able to induce response in imatinib-resistant leukemias [29].

Figure 1. Mechanisms of action of selected antiangiogenic drugs that effect receptor tyrosine kinase pathways.

Another large success of targeted therapy has been the monoclonal antibody rituximab which interacts with the surface antigen CD20 in malignant non-Hodgkin B-cell lymphomas [30], [31]. Rituximab is effective administered alone, but especially when combined with conventional chemotherapy. Recent studies indicate that the mechanism of action is inhibition of the constitutively activated Akt pathways in B-cell lymphomas [30].

These impressive results led to large press headings like the ‘smart bomb’ and ‘sea change’ [32]. Some predicted that cancer soon could be changed to a chronic controllable disease like diabetes, provided the patients continued with an oral non-toxic compound. Such hopes have been the background for the unprecedented search for new targeted drugs and a large number of clinical trials. In the presentation below, the different antivascular agents will be addressed under the heading of the type of cancer in which they are mainly used.

Antiangiogenic therapy

VEGF inhibition

Colorectal cancer

Vascular endothelial growth factor receptors (VEGFR-1, VEFGR-2, VEGFR-3) are specific tyrosine kinase growth factor receptors in endothelial cells. Based on the disclosure of the detailed mechanisms of angiogenesis [4], [33], [34] and lymphangiogenesis [35], [36], a recombinant humanised monoclonal antibody, bevacizumab, was developed as an anticancer agent [37]. Bevacizumab blocks the vascular endothelial growth factor A (VEGF-A). In metastatic colorectal cancer, bevacizumab added to a fluorouracil and folinate regimen (FU/LV) improved response rate (40% with 5 mg/week and 24% with 10 mg/week, vs. 17% for controls), produced longer time to progression (9.0, 7.2, vs. 5.6 months, respectively), and longer survival (21.5, 16.1 vs. 13.8 months) [38]. These data were confirmed in subsequent studies [39], [40]. In a pivotal randomised phase III clinical trial, bevacizumab was added to bolus fluorouracil and folinate plus irinotecan (IFL) [41]. This increased the median duration of survival from 15.6 to 20.3 months, while the secondary endpoint, median duration of progression-free survival increased from 6.2 to 10.6 months. Recently, bevacizumab added to a combination of fluorouracil, folinate and oxaliplatin (FOLFOX4) for patients with previously treated colorectal cancer metastases, improved survival by 2 months (Figure 2) [42]. However, despite the success of this drug, the median survival remains about 20 months for metastatic colorectal cancer [43–45]. Bevacizumab is currently widely used in advanced colorectal cancer, and several randomised phase III trials are ongoing to establish its role in adjuvant combination chemotherapy regimens. Recently, an oral small molecule, inhibiting the VEGFR-2 tyrosine kinase activity, showed clinical efficacy, making this treatment more convenient for the patients [46]. New developments in imaging of angiogenesis clearly document the effect of anti-angiogenic directed therapy in colorectal cancer [47–50].

Figure 2. Prolongation of median overall survival from randomized clinical trials using targeted antiangiogenic substances for different malignancies. Open bars represent variations between different studies. Note that the drugs are given in different settings. Use in 1. line or later will influence the results.

Other cancers

Bevacizumab also exhibited effect in lung cancer, where a median increase in progression-free survival and overall survival of 2 months has been reported by adding bevacizumab (15 mg/m2) to paclitaxel and carboplatin [51–53]. The drug is now FDA approved for lung cancer patients who meet the original eligibility criteria for the studies, which excluded squamous cell histology, brain metastases, and being on anti-coagulants [54]. Also in breast cancer and renal cancer, improved response was seen for bevacizumab added to chemotherapy, but generally the effect is less than that documented in colorectal cancer [55]. Combinations of bevacizumab and new small molecular inhibitors are currently being investigated [56]. VEGF inhibition also exerts pronounced effects on other angioproliferative disorders (neovascular macular degeneration, angioproliferating glaucoma) [57–60]. Bevacizumab has been the leading drug for inhibition of angiogenesis, but a wide range of different drugs are now in early clinical testing [37].

Multiple myeloma

Thalidomide was originally introduced as an anticonvulsant. Later thalidomide became popular as a sedative and antiemetic drug during pregnancy, until its teratogenous effect was discovered in the 1960s [61]. This drug has important immunomodulating properties, and decreases VEGF and beta fibroblast growth factor (bFGF), both known proangiogenic factors. Thalidomide is today standard of care, alone or in combination with chemotherapy, for multiple myeloma, which is the second most common haematological malignancy [61–63]. In controlled studies it delays time to progression with 5 months (from 4 to 9 months). Newer more potent derivatives are currently under investigation. There is, however, substantial toxicity associated with this drug (see below).

EGFR inhibition

Epidermal growth factor receptor (EGFR) belongs to the ErbB family of growth factor receptors expressed on the surface of many cells. The ErbB receptors regulate specific pathways that are crucial for cell growth (proliferation, apoptosis, angiogenesis) [64]. Treatment with EGFR tyrosine kinase inhibitors inhibits angiogenesis by decreasing VEGF [65]. Specific activating molecules (ligands) exist for the different types of receptors. The receptors generally consist of an extracellular growth factor receptor part, a transmembrane domain and an intracellular tyrosine kinase regulatory site (Figure 1). The tyrosine kinase activity can be blocked by specific small molecules, while the growth factors or the external receptor domain can be blocked by specific monoclonal antibodies. The receptors exist as monomers that form dimers when activated. ErbB receptors are up-regulated in many tumour cells. However, EGFR inhibitors have effects in tumour cells expressing less number of receptors per cell than their corresponding normal cells.

Breast cancer

The human epidermal growth factor 2 (HER-2/ErbB2) is a receptor tyrosine kinase overexpressed in 20-30% of human breast cancers, and is associated with poor prognosis [66]. Currently, no specific ligand is associated with HER-2, which tends to form heterodimers with other members of the ErbB family. A humanised monoclonal antibody, trastuzumab, was developed for inactivation of the extracelluar part of HER-2 receptor after intravenous injection [66]. In a phase II study in HER-2 positive, heavily pretreated breast cancer patients with advanced metastases, trastuzumab yielded objective response in 15%. This included eight complete responses and 26 partial responses among 222 patients, and the median duration of response was 9.1 months [67]. In a subsequent study on metastatic disease, trastuzumab alone as first line therapy exhibited an objective response rate of 26% (seven complete responses) in 114 patients [68]. The response was 35% in a subgroup having 3+ or 2+ HER-2 overexpression. Half of the patients had no disease progression at 12 months follow up. These promising results were confirmed in a randomised study where trastuzumab was added to standard chemotherapy (an anthracycline and cyclophosphamide or paclitaxel) vs. chemotherapy alone in 469 patients [69]. There were significantly better results in the group treated with trastuzumab: objective response in 50% vs. 32%; longer time to disease progression, 7.4 vs. 4.6 months, and longer survival, 25.1 vs. 20.3 months. However, the trial also revealed an important difference in side effects as cardiac dysfunction occurred in 27% of the patients treated with trastuzumab in contrast to 8% in the chemotherapy alone group. Based on the effect observed in advanced metastatic disease, trastuzumab was added to paclitaxel after doxorubicin and cyclophosphamide in a randomised adjuvant study, including 133 patients in the trastuzumab arm, and 261 in the control arm [70]. After three years the trastuzumab group had 12% better survival than the control group. Another trial reported the effect of one year trastuzumab administered as an injection every third week after locoregional therapy with surgery and/or radiation therapy, and at least four cycles of neoadjuvant or adjuvant chemotherapy [71]. The initial analysis revealed an absolute difference in disease-free survival of 8.4% after two years. In summary, five studies have used different adjuvant regimens with trastuzumab, and clearly demonstrated a benefit. More impressive, a review of more than 13,000 patients, included in four adjuvant studies, reported that trastuzumab almost halved the risk of recurrence in this population [72–74]. Therefore, adjuvant trastuzumab is current standard of care for HER-2 positive primary tumours after local curative treatment. The new small molecule, lapatinib, is an oral receptor tyrosine kinase inhibitor that blocks both HER-2 and EGFR [75]. Lapatinib plus carboplatin vs. carboplatin alone were used as salvage therapy in a randomised trial in locally advanced or metastatic breast cancer after progression on regimens including an anthracycline, a taxane and trastuzumab [76]. The overall response rate was 22% after combination therapy vs. 14% in the carboplatin group (p = 0.09). The median time to progression was significantly prolonged in the combined group, 8.4 vs. 4.4 months, respectively (P < 0.001), while the survival was similar in both groups.

Lung cancer

Great expectations were raised when a clinical trial of a small molecule EGFR tyrosine kinase inhibitor, gefitinib, 250 mg orally daily, caused objective tumour response in 13.6% for a duration of 8.9 months in non-small cell lung cancer (NSCLC) patients failing two previous chemotherapy combinations, or failing chemotherapy due to toxicity [77]. Then two randomised phase II trials (Ideal 1 and Ideal 2) in NSCLC reported an objective response approaching 20% in second-line patients, and around 10% in those patients having two or more previous chemotherapy regimens [78]. The median survival in the two studies approached 6–8 months. However, two large randomised studies (Intact 1 and Intact 2), including more than 1,000 patients, failed to show an improvement of tumour response or survival when gefitinib was administered as first line therapy in combination with two different chemotherapy regimens [78]. Later, a retrospective molecular analysis of EGFR mutations and amplifications from the above mentioned studies, revealed that the response to gefitinib was related to EGFR mutations, which correlated with previously identified clinical factors like adenocarcinoma histology, absence of smoking history, female sex, and Asian ethnicity [79]. Then a randomised phase III clinical trial reported a significant survival benefit for patients with advanced NSCLC treated by another small molecule inactivating the tyrosine kinase ATP binding site, erlotinib [80]. In a study including 731 patients, the response rate in the erlotinib group was 8.9% vs. <1% in the control group, while the median duration of response was 7.9 vs. 3.7 months, respectively. Overall survival was prolonged by two months. The difference in response in the gefitinib and erlotinib trials may be related to the fact that gefitinib was used at a dose of 42% of the maximum tolerated dose while erlotinib was used at the maximum tolerated dose [81]. A recent trial failed to show any benefit from erlotinib combined with cisplatin and gemcitabine in advanced NSCLC [82]. EGFR inhibition is still considered a promising therapy in aerodigestive tumours, including squamous tumours from the head and neck area [83].

Colorectal cance

EGFR is up-regulated in 60–80% of colorectal cancer tumours [84]. In 57 patients failing an irinotecan containing regimen for metastatic colorectal cancer, 10.5% obtained a partial response on cetuximab, a humanised monoclonal antibody binding to EGFR [85]. The time to tumour progression was only 1.4 months, with a median duration of response of 4.2 months. In another phase II study on 346 patients with metastatic colorectal cancer failing irinotecan, oxaliplatin and fluoropyrimidines, cetuximab achieved similar results with median progression-free survival of 1.4 months and survival of 6.6 months [86]. In a phase III trial patients failing an irinotecan-based regimen, received either cetuximab alone (111 patients) or cetuximab plus the same chemotherapy dose and schedule they previously failed on (218 patients) [84]. The objective response was 10.8% in the cetuximab group, but 22.9% in the combined group. The median time to progression (1.5 vs. 4.1 months) and median survival (6.9 vs. 8.6 months) also favoured the combined group. However, side effects like asthenia, acne-like skin rash, nausea and vomiting were also more frequent in the combined group.

While initially only given to EGFR expressing tumours, the drug seems also to be effective in tumours that do not express EGFR [87]. Patients with tumours expressing high levels of the EGFR ligands epiregulin and amphiregulin seemed more likely to have disease control by cetuximab [88]. Based on the effects in second and third line chemotherapy, cetuximab combined with chemotherapy is now in phase III clinical trials as first line therapy for metastases from colorectal cancer [89], [90]. Another monoclonal antibody directed against EGFR, panitumumab, was administered to 463 patients in a randomised trial including chemotherapy-refractory metastatic colorectal cancer as best supportive care alone vs. best supportive care and panitumumab [91]. Despite a significant difference in median progression-free survival, the difference was only 7.3 vs. 8 weeks (5 days). Recently the effect of monoclonal antibodies against EGFR seems to depend on wild type of KRAS as mutations in this gene predict resistance [92–94].

In pancreatic cancer, 569 patients were randomised between standard gemcitabine and erlotinib and gemcitabine [95]. The overall survival was statistically significant longer in the combined group (P = 0.038), but the difference in median overall survival was hardly of clinical significance: 5.91 months in the gemcitabine group and 6.24 months in the combined group.

Renal cell carcinoma

Renal cell carcinomas are refractory to most standard chemotherapy regimens. It is well known that primary tumours and metastases from renal cell carcinomas generally are well vascularised, often causing bleedings during surgical resections. The hypoxia-inducible factor 1α (HIF-1α) is an important regulator of angiogenesis [96]. New insights into the molecular mechanisms in renal cell carcinomas reveal that for clear cell carcinomas there is accumulation of HIF-1α due to a defect in the von Hippel-Lindau protein which normally addresses HIF-1α for degradation (ubiquitinaton) [97], [98]. HIF-1α then causes overexpression of VEGF, transforming growth factor α (TGF-α, a ligand for EGFR), and platelet derived growth factor chain β (PDGF-β) which also stimulates angiogenesis [99]. On this background, bevacizumab was tested in metastatic renal cell cancer in a randomised phase II trial [100]. A dose of 10 mg/kg improved progression-free survival by 2–3 months compared to placebo, but there was no survival benefit. Then an oral, multi-targeted tyrosine kinase inhibitor, sunitinib, targeting VEGFR type 1-3 and PDGFR (α and β), proved to be effective in metastatic clear cell renal cancer in a pooled analysis of two phase II studies [101]. The drug induced partial response in 45% of the patients (36% when assessed by an independent review committee), and the median progression-free survival was 8.1 months. In a randomised phase III study including 750 patients, the same regimen of sunitinib (50 mg orally each day) was compared with interferon α which can be considered standard therapy [102]. In this study the sunitinib arm yielded about 20% objective responses and a median progression free survival of 11 months, in contrast to interferon α where the corresponding endpoint was 5 months. It is also important that the patients on sunitinib reported significantly better quality of life. Others used a combination of bevacizumab (10 mg/kg daily) and erlotinib (150 mg orally daily) for metastatic renal cell cancer to attack different targets (VEGF, EGFR, respectively) in 59 evaluable patients [103]. The combination was well tolerated and resulted in 25% partial responses and a median progression free survival of 11 months. Sorafenib, another tyrosine kinase inhibitor, targets both tumour cells and tumour vasculature as it inhibits VEGFR type 2, B-Raf, PDGFR, Fms-like tyrosine kinase (Flt-3) and stem cell factor receptor (c-Kit) [104]. When sorafenib was given orally twice daily for 12 weeks, 73 out of 202 patients with metastatic renal cell cancer obtained tumour shrinkage of > 25% [104]. The responding patients were then randomised to continued therapy or placebo, which resulted in a median progression-free survival of 24 weeks in the treated group vs. 6 weeks for the placebo group. Then a follow-up randomised phase III study including 903 patients showed a median progression-free survival of 5.5 months in the sorafenib group against 2.8 months in the placebo group [105]. Thus, the multi-targeted tyrosine kinase inhibitors yield currently the best available therapy in clear cell renal cancer.

Matrix metalloproteinase inhibitors

Matrix metalloproteinases (MMP) are a family of proteolytic enzymes with effects on various components of the extracellular matrix, where the effect of MMPs are controlled by endogenous tissue inhibitors of MMPs (TIMPs) [106]. This is important for vascular remodelling and angiogenesis, and MMPs are partly regulated by VEGF [107]. The role of MMP inhibitors is currently unclear as randomised phase III trials have demonstrated negative effect on survival [108].

Vascular disrupting agents

The vascular targeting drugs selectively destruct the vasculature of malignant tumours, thereby causing rapid and selective shutdown of blood supply and thus indirect killing of the tumour cells [109–111]. The concept was introduced in the 1980's by Denekamp who underlined that the tumour vascular endothelial cells are proliferating, and that the undifferentiated vessels differ from corresponding normal ones, making them particularly vulnerable for vascular targeting drugs [112], [113]. In animal studies, the Vinka alkaloids vincristine and vinblastine, the microtubule disrupting agent combretastatine A4, arsenic trioxide, flavone acetic acid, tumour necrosis factor and 5,6-dimethylxanthenone-4-acetic acid are the most studied drugs [114–123]. These drugs are now in clinical phase II studies after showing some objective effect in phase I studies [110], [124]. It is too early to judge their true clinical role as side effects like myocardial infarction, tumour bleeding, tumour pain, as well as fatigue and nausea are associated with their use. However, these drugs are of particular interest as their anti-tumour effects are enhanced by hyperthermia [75], [116], [119], [120].

Low dose or metronomic chemotherapy

While cytotoxic chemotherapy generally is administered as courses at highest tolerable doses for some days, with rest intervals of 2 to 3 weeks, metronomic chemotherapy is administered as regularly repeated (i.e. metronomic) low, nontoxic doses, aiming at destruction of endothelial cells in the irregular, newly formed tumour vessels [125–129]. The aim of the therapy is therefore indirect tumour cell killing by cutting off the supply of nutrition and oxygen, necessary for maintenance of the tumour cell population. This dosage strategy not only works well in experimental animal tumours, but also show objective tumour regressions in advanced human malignancies, and even in some tumours generally considered resistant to standard chemotherapy [130–132]. One of the known mechanisms of metronomic therapy is induction of gene and protein expression of thrombospondin 1, a potent inhibitor of angiogenesis [126], [127]. A new aspect of metronomic dosing is the reduction of cells with stem cell characteristics in experimental glioma xenografts [133]. Even more intriguing is the reduction of circulating endothelial stem cells after metronomic dosed chemotherapy in animal models [107], [134–136]. However, the contribution of bone marrow derived endothelial cells to the tumour vasculature seems smaller in man than in rodents [137]. As there are reduced side effects with metronomic dosed therapy, it seems particularly well suited for palliative treatment.

Hyperthermia as an antivascular agent

The biological rationale for hyperthermia as a method for increasing the effect of radiation and chemotherapy is well documented [138–141]. From experimental and clinical studies it is well known that hyperthermia has effects on blood flow, causing vascular shut down at high temperatures, but also a positive effect on tumour tissue oxygenation at mild temperatures [122], [142–146]. It has been clearly demonstrated that hyperthermia exhibits antiangiogenic effects, and proliferating endothelial cells in vessels from malignant tumours seem more sensitive to heat than the microvessels of normal tissues [116] [147–149]. Possible mechanisms may be changes in HIF-1α, VEGF or enhancement of the tumour necrosis factor [142], [150–152], but the exact mechanisms of action remains unclear. The known positive interaction with vascular disrupting drugs indicates that hyperthermia also influences some crucial mechanisms for endothelial cell proliferation and maintenance.

The important progress of hyperthermia in the clinic is well documented in randomised controlled studies. When hyperthermia was given with adequate equipment it enhanced the effect of radiation in localised head and neck lymph node metastases [153], recurrent cutaneous malignant melanomas [154], [155], recurrent breast cancer [156], [157], soft tissue sarcomas [158–160] and localised brain tumours [161]. Of pelvic malignancies, particularly the effect in cervical cancer is remarkable [162–164]. The effect of radiation and heat seems even further improved by adding cisplatin, and two controlled studies are now ongoing testing this trimodality treatment in cervical cancer [165]. Combined with chemotherapy, hyperthermia has significantly enhanced the effect observed in superficial bladder cancer [166] and recently for locally advanced sarcomas [167]. It is outside the scope of this paper to go into detail on the clinical results. However, it can be concluded that hyperthermia can increase the control rate of radiation alone by about 50% (i.e. from 40% to 60%), and the effects observed in primary situations tend to be durable, i.e. the effect is not only a transient phenomenon.

Hyperthermia may also be used as a method to direct drugs towards the local tumour, both for nanoparticles and liposomes [168–173]. Recently, several groups have used the local effects to selectively deliver gene therapy to tumours by activation of a heat responsive promoter construct [174–177]. Hyperthermia has many mechanisms of action, but hyperthermia deserves also to be considered as one of the most clinically established antivascular treatments.

Toxicity issues

Terms like smart bombs and targeted therapy imply that these drugs find their targets selectively, thus avoiding the side effects tightly associated with classic anticancer agents, often termed cytotoxic agents. Due to the presumed selective effect of ‘designer’ drugs, the established maximum tolerated dose was frequently claimed to be irrelevant. However, as large number of patients have been exposed to the drugs, it has become clear that apart from some drug-specific side effects, many of the general side effects (nausea and vomiting, diarrhoea, allergic reactions and bone marrow suppression) are also a problem with targeted drugs (Table 1) [107], [178–180]. This may simply reflect the disturbance on growth factor signalling pathways involved in many normal processes [181]. It has especially been focused on dermatological side effects for EGFR inhibiting agents [182] and cardiotoxic effects of trastuzumab and imatinib [183–185]. For anti-angiogenic agents, effects on wound healing are relevant when surgical treatment is planned or must be performed during therapy [186], and endothelial effects may also initiate thromboembolic events [56], [181], [187]. For bevacizumab, hypertension, proteinuria and bleeding should be carefully addressed [188]. Thalidomide is associated with substantial toxicity: thromboembolism which can be controlled by anticoagulants, somnolence, neurotoxicity, obstipation, and skin rash [61–63]. The main acute side effects from hyperthermia are general discomfort and some localised pain during the treatment sessions. Serious late complications after local and regional heating, apart from some local blistering, have generally not reached statistical significance above that of radiation and chemotherapy alone. In a study of advanced rectal cancer, addition of hyperthermia to radiation therapy had no detrimental effect on the patient's quality of life [189]. However, severe neurological side effects and tissue necrosis have been described in some adult patients after regional heating [190], [191], and avascular femoral osteonecrosis appeared in 14% of children less than 5 years after multiple hyperthermia sessions [192]. The neurological damage is probably determined by the sensitivity of the nerve vasculature which should be kept under a dose of 30 min at 44°C or equivalent. Hopefully, the introduction of specific absorption rate (SAR) based treatment planning and continuous non-invasive monitoring, will help in delivering homogenous heating and reduce thermal hot spots [193–197].

Table 1.  Some of the side effects recently associated with use of antiangiogenic drugs.

Side effects	Typical for
• Myocardial infarction	Vascular disrupting drugs, trastuzumab
• Hypertension	Bevacizumab
• Reduced wound healing, fistulae	Bevacizumab
• Diarrhea	EGFR inhibitors
• Skin toxicity	EGFR inhibitors, hyperthermia
• Fatigue	Common
• Allergy	Common
• Nausea	Common
• Bone marrow depression	Several
• Hypomagnesemia	Bevacizumab
• Teratogenicity	Thalidomide
• Bleeding	Vascular disrupting drugs, bevacizumab
• Neurotoxicity	Thalidomide, hyperthermia

Discussion and conclusions

The major curative treatment modalities for cancer remain surgery and/or radiation therapy. For some tumour types (malignant lymphomas, testicular germ cell tumours and leukemias) chemotherapy may be a curative treatment. For the majority of patients treated with chemotherapy, the goal of the therapy is palliation and extension of lifetime, except in adjuvant settings where the addition of chemotherapy to standard locoregional therapy increases cure rate (often in the range of 10–15%). Hyperthermia remains the most reliable sensitiser to increase the efficacy of radiation and chemotherapy [141], [198]. The tumour response tends to be durable as evidenced by parallel survival curves in the hyperthermia group and the control group. For most targeted therapy, with important exceptions for imatinib in chronic myelogenous leukemias and GIST, rituximab in malignant lymphomas, and trastuzumab in breast cancer, the effects are most often palliative for a few months duration. Bevacizumab improves the survival of metastatic colorectal cancer by several months, but its role combined with chemotherapy as adjuvant therapy after surgery is under investigation. Antiangiogenic drugs are expensive, but both trastuzumab and cetuximab are considered cost effective [199], [200]. However, the investment in hyperthermic equipment and staff to run the facilities are also cost effective [163]. These two treatment options should both be further evaluated in the clinic. Combinations of antiangiogenic drugs and hyperthermia may prove effective both experimentally and in the clinic. Far more money is presently invested in potential new drugs, instead of letting patients with suitable tumours gain access to hyperthermia for conditions where phase III studies have documented effects. Certainly, many new drugs will give us better treatment options in the years to come. Nevertheless, hyperthermia is already a valuable option for selected malignancies, and the machines for administering heating are today much more reliable than those used 30 years ago. Hyperthermia must therefore get the necessary funding to become available in major cancer centres.