Metronomic anti-cancer therapy – an ongoing treatment option for advanced cancer patients

The therapeutic concept of administering agents (cytotoxic/-static,non-cytotoxic and/or targeted drugs) continuously at lower doses relative to MTDs in the case of cytostatic and cytotoxic drugs or continuously at tolerable doses as in the case of targeted drugs without drug-free breaks over extended periods known as ‘metronomic therapy’ (MT), is increasingly being recognized as an experimental option for treating cancer. In comparison with MTD-defined chemotherapy (CHT) regimens, metronomic therapy has demonstrated reduced toxicity. More importantly, several phase II trials have shown that metronomic therapies showed anti-cancer activity in different cancer types with different drugs. The mechanistic basis of metronomic therapy using cytotoxic/static drugs is believed to be primarily anti-angiogenic, either by direct killing or inhibiting endothelial cells (ECs) in the tumor vasculature, killing bone-marrow-derived endothelial progenitor cells, stimulating the immune system, directly affecting tumor cells through a drug-driven effect as well as specificly inhibiting a target when targeting drugs were used in additional to metronomic therapy. The induction of senescence in cancer is another possible explanation for the principles behind ‘metronomic therapy’.


Introduction
The concept of using anti-angiogenic approaches as an anti-cancer treatment strategy was suggested by Judah Folkman in 1971 [1]. Three decades later the first approval of a 'new' anti-angiogenic agent was attained [2], although the debate of how to use bevacizumab optimally in which patients is still ongoing. Several physicians argue that using anti-angiogenic agents such as bevacizumab continuously and not only as first line does not appear to be a realistic option due to the high costs involved. Many conventional cytotoxic drugs have significant antiangiogenic effects [3]. Conventional chemotherapy CHT is normally administered near the maximal tolerated dose (MTD), typically in 3-week cycles. When CHT is used in this way with a 3 or more weeks drug-free period, the host can recover from adverse effects and the vascular damage will rapidly be repaired whereas cancer cells have less chance of recovering. Nevertheless during the drug-free period, regrowth of parts of the cancer cells occurs, especially if the drug-free period is too long. It has been shown, that by shortening the drug-free period between CHTs, the anti-angiogenic effects can be largely augmented [4] Unfortunately, the study of Browder et al., was carried out with rapidly proliferating tumors after subcutaneous transplantation in mice, but the use of slowgrowing, spontaneous tumors would be a better method to evaluate this topic. This as metronomic dubbed CHT can be optimized when combined with targeted agents [4]. A metronomic CHT is characterized by the following principles (see Table 1).

Table 1. Characteristics of metronomic chemotherapy
• frequent (dose-dense) CHT administration without any interruptions • not using the maximal tolerated dose (MTD) include a biological optimized dose (BOD) • no application of hematopoetic growth factors • preference for oral drugs • low incidence of treatment related side effects • potential for delayed development of resistance This approach can be further refined. In fact, the only available anti-neoplastic drugs which can be administered orally are limited to: cyclophosphamide, methotrexate, trofosfamide, capecitabine, S-1, UFT, navelbine, etoposid, temozolomide, dexamethasone, estramustin, chlorambucil,

Anti-angiogenic properties
Growth of tumor vasculature depends on angiogenesis which is mediated by local sprouting of rapidly dividing endothelial cells (ECs) from pre-existing capillaries and vasculogenesis which requires circulating endothelial precursor cells (CEPs) (angioblast-like) from bone marrow [7]. Intratumoral vascular ECs proliferate rapidly in contrast to the ECs of quiescent mature blood vessels of normal adult tissues. These rapidly proliferating ECs are vulnerable to cytotoxic/static agents (Figure 2).
That anti-tumor agents have endothelial toxicity is not a very new observation [8,9]. A drug that requires a higher concentration to kill endothelial cells than is required to kill cancer cells should not be considered anti-angiogenic. The strength of the available data on which this claim is founded (anti-angiogenic properties of metronomic chemotherapy) varies dramatically. Is every cytotoxic agent really antiangiogenic? Most authors do not draw this conclusion. Indeed, multiple potential criteria are necessary (see Table 2) for defining chemotherapy with anticancer agents Figure 1. Shows typical 'metronomic' chemotherapies in comparison to the standard procedure (repetitive infusions or boli near resp. at the MTD) from [6]. From above to below: 1 st row figures a classical q3w schedule at the MTD; 2 nd row shows a continuous infusion at the MTD; 3 rd row presents a metronomic (daily) oral therapy; 4 th row displays a continuous infusion together with a q3w application of iv chemotherapy (e.g. ECF); 5 th row depicts a q3w iv chemotherapy combined with a metronomic low -dose concept (e.g. XELOX) 6 th row shows a weekly chemotherapy (e.g. Paclitaxel) Figure 2. Metronomic chemotherapy reduces the rapid proliferating angioblast-like CEPs (circulating and growing out from bone marrow) and possibly inhibits proliferating ECs (most of the ECs of `old` vessels do not proliferate). (modified from [126]  Although we might think of anti-angiogenic activity as being mediated by direct effects on endothelial cells, it is equally possible for an anti-angiogenic activity to be mediated by cancer cells. If cancer cells produce proangiogenic factors in an autocrine fashion, which is the driving force behind angiogenesis, then killing cancer cells will effect endothelial cells just by eliminating these survival factors critical to endothelial cells. It would be trivial to postulate that every successful cytotoxic therapy is anti-angiogenic because in each NED (no evidence of disease) resp. CR (complete remission) stage the tumor has diappeared or not visible, but also the tumor`s vasculature now is absent resp. is no longer visible anymore. Greater awareness of the collateral damaging effects of most anticancer therapeutics on the micro-environment of cancer cells should indeed make it possible to optimise this beneficial therapeutic side effect to minimise host toxicity (Figure 3).  Combination therapies of metronomic chemotherapy plus antiangiogenic drugs have shown some anti-angiogenic effects [10,11], demonstrated with the available techniques from 10 to 20 years ago. The effects of cytostatic agents on endothelial cells have been described [12] and in general it is accepted that chemotherapy has, in a schedule dependant way, measurable effects on proliferating endothelial cells but the fracton of endothelial cells which is actually proliferating in the tumors vasculature is low (only 3%). Current knowledge based on, preclinical in vitro and in vivo data as well as emerging clinical data, suggests an anti-angiogenic activity for several existing, commonly used drugs.

Activation of immunity
It is well known that cytotoxic drugs, like anthracyclines, taxanes and especially cyclophosphamide (see figure 3) display important immunostimulatory properties [13,128]. The effects on regulatory T cells (T reg ) seem to be relevant in the context of metronomic treatments. T reg are CD4+, CD 25+ lymphocytes which can inhibit an anti-tumor immune response by suppressing the activity of tumorspecific and tumorunspecific effector cells. An increase of T reg cells can be correlated with tumor progression and lack of treatment response [14]. Several studies have shown that low-dose cyclophosphamide decreases the number of T reg cells, inhibiting their suppressive function [15,16]. Another immunestimulatory effect has been demonstrated, i.e., the induction of dendritic cell maturation [17]. The promotion of dendritic cell maturation with metronomic chemotherapy in a non-toxic concentration was please observed noted and is bridging the gap between metronomic therapy doi: 10.7243/2049-7962-1-32 and immunity (Figure 4). Furthermore, metronomic cyclophosphamide showed a large increase in tumorassocciated NK cells, dendritic cells, and macrophages and occurred after the onset of tumor regression in an immunocompetent, the syngeneic mouse model [128].

Induction of senescence
Lower levels of damage by chemotherapy may trigger senescence associated with antiproliferative responses without activating the cascades of caspase activity that induce cellular apoptosis. For example dosing of 250 nM doxorubicin generates apoptosis (in prostata cells), whereas 25 nM induces senescence [18]. Many senescence-inducing drugs generate DNA damage to produce single-and doublestrand breaks. The induction of senescence in tumors can in fact be achieved by permanently administered, lowdose schedules of cytostatics [19]. Metronomic therapies have no great capacity of inducing apoptosis, but due to their low drug levels, senescence can be one mechanism which occurs during MT. This obsrvations warrants further investigation.

Clinical trials with 'metronomic scheduling' of cytotoxic/-static drugs
According to figure 1 a metronomic chemotherapy is defined as a therapy where at each timepoint a biologically active drug concentration acts. In this scenario certain common chemotherapy protocols include metronomic components. E.g. ECF for gastric cancer uses a continuous infusion of 5-fluorouracil, the maintenance therapy of ALL contains continuous 6-mercaptopurine, even the now commonly used weekly paclitaxel therapies (instead of q3w schedules) have a `metronomic` character as well as the Stanford V chemotherapy regimen for the treatment of Hodgkins disease with different drugs given weekly over three months in lower dosages than in the common ABVD-MOPP or BEACOPP regimens. Even the classical CMF scheme for breast cancer patients contains a `metronomic` part with oral cyclophosphamide over 14 days albeit with a pause of one week between the treatment cycles. The original CMF scheme developed by Bonadonna with orally administered cyclophosphamide (100 mg/m² daily for 24 days) seems to be more effective than the i.v. scheme (600 mg/m² at day 1) which was introduced to simplify the drug schedule and to make it more convenient for patients as shown in a phase III study conducted by the EORTC [50]. The total drug amount in the original treatment protocol was more than doubled in the case of cyclophosphamide. In this case, the `metronomic` schedule reflects a greater dose intensity and is a MTD-defined metronomic chemotherapy CHT. In a recent published phase III study [67] it was shown in gestational trophoblastic neoplasia (GTN) that a bi-weekly i.v. administration of dactinomycin is better than a weekly im injection of methotrexate showing that this pulsed biweekly schedule can be viewed as a metronomic CHT. In the majority of clinical trials, which are titled metronomic CHT, cyclophosphamide (CYC) has an essential part because of its different effects mentioned peviously.. All these classical `metronomic` CHTs chemotherapies use the maximal tolerable dose (MTD) and not a biological optimal dose (BOD). The dosing at the MTD is strongly related to the well known toxicity and therefore a metronomic chemotherapy can be MTD driven, but it can also be dosed far below the MTD, nearly at non-toxic dose levels, a dosing which is not well developed up to now because we use anticancer agents mostly at the maximum tolerated dose.

Clinical studies with metronomic chemotherapy
An increasing number of clinical studies using metronomic therapy schedules have been published in the last years. For this article a thorough literature research was performed in PubMed and Google scholar. 80 publications were found and acted as basis for the discussion in this review. Solid tumors as well as hematologic malignancies were included in this survey. The main focus was on patients with advanced, multiple metastasized a multiple pretreated solid tumors who had progressive disease and sometimes stable disease when entering the metronomic therapy. In the cases with stable disease, the goal of metronomic therapy is to maintain this status.. Therefore the entry criteria were not stringent e.g. always a progressive period before study entry. Three different metronomic schedules were of special interest: 1. metronomic CHT with cytostatics/cytotoxics, 2. the combination of cytostatics with other agents that may have anti-angiogenic effects, 3. the combination of cytostatics with targeted therapies. Besides, a short overview for metronomic therapy in the adjuvant setting was also considered interesting, albeit here the number of studies is limited [96][97][98][99].
The most commonly examinated tumor type is breast cancer. Here we found 16 studies with metronomic therapy, followed by colorectal cancer, glioblastoma, melanoma, non small cell lung cancer, prostate cancer and different histologic types of sarcomas . But results in other tumors with a smaller number of publications such as head and neck cancer, adrenocortical carcinoma, renal cell carcinoma, neuroendocrine carcinoma and ovarian cancer have also been described . Besides, there are increasing numbers of clinical studies who are engaged in hematologic malignancies, like different types of lymphomas including multiple myeloma. Also for children with advanced cancer and glioblastoma metronomic approaches are published. A summary of studies including indication, dosing schedule, response and description of the most special toxicities is depicted in table 3 (here is the literature cited).
Regarding the metronomic schedules for different cancer types, oral CHT drugs such as cyclophosphamide, methotrexate, trofosfamide and etoposide are the main chemotherapeutics used in these trials. Capecitabine and oral 5-FU prodrugs like UFT are also suitable agents and doi: 10.7243/2049-7962-1-32  can be combined with other cytotoxic drugs as an oral formulation for metronomic CHT.
We found a very interesting study with capecitabine for patients with advanced breast cancer [56]. The study included 60 patients, 25% heavily pretreated with three or more prior chemotherapies and 22% had capecitabine previously as palliative treatment with intermittand capecitabine 2000 mg/m² on days 1-14 every three weeks and vinorelbine 60 mg/m² day 1+8 every three weeks. The overall response rate was 24%, a disease stabilization was documented in 62%. Also the 13 patients who were previously treated with intermittend capecitabine had substantial benefit from the metronomic therapy with 1 CR, 1 PR and 7 SD. This is a study demonstrating safety and efficacy of metronomic capecitabine in the palliative setting with a special focus on the quality of life in advanced breast cancer. Cyclophosphamide is predominantly used for metronomic concepts in breast cancer, prostate cancer and lymphoma, often combined with methotrexate. Nearly all studies combine metronomic CHT with an agent that interacts with a specific target. In our survey we found combinations with oral drugs such as thalidomide, letrozole, megestrole acetate, dexamethasone, prednisone and cyclooxygenase-2-inhibitors, e.g. celecoxib, rofecoxib and etoricoxib as well as the PPARγ-agonist pioglitazone. Dalteparin administered subcutaneously was combined with low dose cyclophosphamide, MTX and prednisone for patients with advanced breast cancer. Interferon-α was combined with etoricoxib, pioglitazone and capecitabine in advanced renal cell carcinoma. The underlying rationale is to enhance the therapeutic effect of the low dose/dense cytotoxic regimens. Furthermore, known cytotoxic/-static drugs which are only available for intravenous application are of current interest. In colorectal cancer patients drugs such as irinotecan were given in a metronomic fashion as low dose continuous infusion for 3 weeks (1,4 to 4,2 mg/d) followed by a 1 week rest [29]. The treatment was well tolerated without toxicity and resulted in a disease stabilization of 20%. The combination of low-dose irinotecan in a weekly schedule with S1 showed remarkable results with an overall response rate of 43,5% and a disease stabilization of 37,5% with low toxicity. For example only 1 Grade 3 diarrhea being observed. Other intravenous applicable drugs suitable for metronomic scheduling are paclitaxel given as a weekly infusion, in this case used for patients with advanced head and neck [35], breast [49] and ovarian cancer [66]. Vinorelbine is applied weekly for patients with advanced breast cancer. Weekly docetaxel is more favourable in combination with daily trofosfamide for advanced non small cell lung cancer although docetaxel in a weekly i.v. schedule was inferior in breast cancer and prostate cancer in comparison to a q3w schedule. Improvement of a metronomic CHT in terms of tumor stabilization or prolonged progression free survival may be reached by the combination of one or two oral cytotoxic drugs with an anti-angiogenic drug such as an anti-VEGFR-2 antibody or a small molecule multi-targeted VEGFR-2 antagonist receptor tyrosine kinase inhibitor. A summary of studies is depicted in table 4.
Two studies examined be-weekly bevacizumab in combination with low dose oral cytostatic agents for breast cancer, in three studies this schedule was used in ovarian cancer. Imatinib in combination with conventional antitumor drugs was investigated in patients with hepatocellular carcinoma [90]. A combination with lapatinib was found in a case report of a patient with advanced breast cancer [91]. Weekly rituximab was combined with four oral antiproliferative oral drugs as metronomic treatment for mantle cell lymphoma [92]. Two clinical trials evaluated the effectiveness of sorafenib when combined with a low dose cytotoxic agent [93,94]. The indications were hepatocellular and renal cell carcinoma. Trastuzumab shows antiangiogenic effects in a metronomic schedule for advanced breast cancer patients when combined with cyclophosphamide and methotrexate [95].
Interestingly, there are three clinical studies which examined the effectiveness of metronomic therapy in the adjuvant setting, and one study which examined metronomic therapy in the neoadjuvant setting ( Table 5) has been reported.
The first one was performed in patients with colorectal cancer: CPT-11 was given 3 times weekly, followed by one week rest. The combined drug for the metronomic schedule was UFT given daily, 6 month.. in combination with CPT-11 afterwards followed by six months daily as monotherapy. In another trial, patients with glioblastoma received temozolomide after standard radiochemotherapy in two different schedules: a dose dense and metronomic regimen until tumor progression. The third study examined the effectiveness of vaccination therapy combined with external beam radiation in prostate cancer patients. The urgent demand of large randomized studies is obvious otherwise no acceptance of such a treatment strategy can be expected.
In the study with neoadjuvant therapy for breast cancer the conventional schedule with doxorubicin and cyclophosphamide every three weeks was compared with the weekly application of these drugs and G-CSF support. The primary outcome was microscopic pathologic complete response (pCR) at surgery.
Analyzing all studies depicted in table 1-3, the number of patients included in a study varies from 1 to 631. Fifty-five studies included only ≤ 50 patients. We defined responders as having either complete remission, partial remission or stable disease. This is different from normal, when responders are only patients with partial and complete response. The judgement NC (no change) can be a minor regression (less than 30%) or a small progression (less than 20%). The response is described as the absolute number out of the total number of patients evaluated. The results are as follows: in 74 trials, patients responded to therapy. Only one study resulted in progressive disease. 2 twoarm studies resulted in progressive disease and response in either of the treatment arms. The first was a study for patients with advanced melanoma. Treatment arm A consisted of trofosfamide 3x50 mg/d continuously, arm B included the same regimen plus rofecoxib 25 mg/d and pioglitazone 60 mg/d [56]. Response was documented in the combination arm as benefit in progression free survival. The second two-arm study was performed with patients with advanced prostate cancer. Treatment consisted of docetaxel in escalating doses and zoledronic acid (2 mg fixed dose), every two weeks in 2 different sequences [74]. Treatment with zoledronic acid upfront followed by docetaxel showed no response.
A study evaluating the effectiveness of metronomic therapy in the adjuvant setting in patients with colorectal cancer resulted in non-inferiority for 5-year overall survival. The adjuvant treatment with temozolomide for patients with glioblastoma consisted of two arms: temozolomide administered as a dose dense and a metronomic schedule.
Here the results were in favour of the dose dense treatment arm. For prostate cancer patients the adjuvant treatment consisted of vaccination therapy with interleukin 2 before and after definitive external beam radiation therapy and resulted in the induction of prostate-specific immune responses.
Neoadjuvant treatment of inflammatory or locally advanced breast cancer with a metronomic therapy showed -in comparison to the conventional schedule -a higher pCR-rate. Further studies with a metronomic approach doi: 10.7243/2049-7962-1-32

Toxicity of metronomic chemotherapy
The patient populations, drug combinations as well as tumor types investigated thus far are very heterogenous as are the efficacy data which were reported. In total, metronomic CHT alone or in combination demonstrated a good tolerability when daily given, so this is obligatory. Most common side effects were grade 1 nausea and grade 1 to 2 anemia and neutropenia as well as grade 1 to 2 fatigue. But a severe problem is the absence of large phase III studies comparing CHT against metronomic CHT. The number of treated patients with sampled toxicity data is low because phase II studies include only a small number of patients in general far below a hundred patients. Overall, metronomic CHT has often been described with minimal toxicity offering a significant benefit for the patients including quality of life. Theoretically high cumulation over time of etoposide, temozolomide and cyclophosphamide can lead to secondary leukemia, myelodysplastic syndromes (MDS) or resistance. Furthermore, if anti-angiogenic drugs are added to a metronomic CHT, it is unclear how to dose these drugs. All approved anti-angiogenic drugs have a distinct side effect pattern, some of them have an even broader spectrum of unpleasant effects. Metronomic CHT without major side effects can easily become toxic by the addition of modern targeted drugs. Especially the difficulty to treat a common toxicity like fatigue can become a problem, as well as gastrointestinal symptoms e.g. nausea, vomiting and diarrhea. Antagonists for acute toxicity are available, but treating chronic symptoms may be much more difficult. The potential advantage of anticancer therapies with less side effects than the common MTD-driven CHT turns to a disadvantage when a toxicity is daily present. The main toxicities Grade 3 and 4 described in all studies are depicted in table 1-3. We selected the most important side effects according to their relevance in clinical practice. Although metronomic treatment concepts use low doses of cytostatics, this therapy is not without side effects ( Table 6). In 21 studies neutropenia Grade 3+4 was documented. In 9 studies elevated liver enzymes were seen. Eight studies documented lymphopenia, 17 showed anemia, thrombocytopenia was seen in 19 and fatigue in 10 trials. In summary, 5491 patients out of all studies were treated. Regarding the absolute number of patients experiencing toxicity, the most frequent side effect is neutropenia in 1122 patients (20,6%). Anemia is the second most frequently listed side effect, documented in 518 patients (9,5%) but was never an indication for an interruption or termination of a metronomic therapy. Elevation of liver enzymes was seen in 48 patients, that is 0,9% of all patients assessed in the studies for this review.

Metronomic therapies with non-cytotoxic/-static drugs
Many compounds and drugs which are already approved and on the market, but not listed for anti-cancer treatment, were screened for their anti-angiogenic and anti-cancer properties and there are some interesting drugs for combinations in metronomic schedules. COX-2 is the inducible form of the enzyme that metabolizes the lipid arachidonic acid to prostaglandin H2, which is the first step of prostaglandin production. These compounds are widely in use as analgetic and anti-inflammatory drugs. The anti-angiogenic activity of COX-2 inhibitors is well known (100). It has been shown that the additional effects of the COX-2 inhibitors on apoptosis were distinct from their effects on COX-2 inhibition [101], but both effects are very welcome for an anti-cancer treatment approach. In general, COX-2 is up-regulated during development of cancer from early hyperplasia to metastatic cancer, especially in colon cancer, but in other cancer types as well. COX-2 was consistently and more intensively observed in metastatic lesions compared with the primary tumor. COX-2 was detected in non-cancerous cells immediately adjacent to tumor cells and in the angiogenic vasculature of the tumor, but not in the vasculature of normal tissues [102]. Furthermore, COX-2 inhibitors like celecoxib increase the intracellular accumulation of doxorubicin and enhance the cytotoxic effects via NF-κB [103]. In the adenoma prevention with celexoxib (APC) study, it was shown that celexocib can reduce the development of adenomas in a dose dependent manner [104]. It is a long known und recently updated fact that the daily intake (a metronomic  schedule) of low-dose aspirin can prevent colon cancer [105]. Another interesting and novel drug for treating or preventing malignant diseases which is coming into focus is metformin. It is widely used in the treatment of diabetes mellitus type 2 to reduce insulin resistance. Population studies have suggested that metformin used in diabetic patients decreases cancer incidence and mortality. In a recent retrospective study it was found that breast cancer patients with diabetes receiving metformin have a significant higher pCR rate after neoadjuvant chemotherapy than patients not receiving metformin [106]. The mechanism of this effect is a matter of discussion and ongoing studies. Two important pathways are being discussed that are involved in cancer growth. One is the insulin/insulin-like growth factor-1 (IGF-1) signaling pathway which is activated when nutritients are available; the other is the adenosine mono-phosphate-acitvated protein kinase (AMPK) pathway, activated when cells are starved for carbohydrates [107]. Furthermore, low-dose metformin in combination with cyctostatic/-toxic agents like doxorubicin is able to kill cancer stem cells [108] and metformin is able to inhibit mTOR [109] as this is the effect of the activation of the AMP-activated protein kinase (AMPK) induced by metformin. The facts about this anti-diabetic drug related to cancer control have been recently reviewed and summarized [110]. In our opinion this topic is a very exciting one and has led to different clinical trials, including translational research to explore this very old and inexpensive drug in the oncology world. Metformin use is associated with a better survival of diabetic patients with pancreatic cancer [129]. The question is whether metformin treatment or the diabetes mellitus itself, with all the metabolic changes, is the explanation.
Many up-stream signals like epidermal growth factor receptor (EGF-R) or insulin-like growth factor receptor (IGF-R) feed into the PI3-kinase/Akt cascade ( Figure 5).
The radiosensitizing effect of HIV protease inhibitors has been linked to the inhibition of phospho-Akt. A case report has been published in 1999 with a complete response of a Kaposi sarcoma which had been treated with an antiretroviral therapy [111]. Nelfinavir is a lead HIV protease inhibitor which inhibits AKT. It plays a crucial role in many pathways of growth factors and stimulates cancer growth. It would be of utmost interest to have an AKT inhibitor available just to interfere with these different pathways. The drug further induces a G 1 cell arrest and inhibits the proliferation of cancer cells in the NCI60 cell line panel [112]. Noteworthy, prolonged exposure rather than C max are more important for tumor inhibition, suggesting a 'metronomic' schedule of this drug. One phase I study has been published with cisplatin, gemcitabine and oral nelfinavir [113]. Because there are abundant safety data on HIV protease inhibitors accumulated in thousands of patients, such a drug can easily be added to metronomic concepts introducing a new mechanism of action to such a concept [114]. The thiazolidinediones represent another class of drugs featuring anti-neoplastic effects as peroxisome proliferatoractivated receptor γ agonists. As cartooned in a review [115], such drugs decreased the intracellular adhesion molecule-1, the vascular cell adhesion molecule-1, increased nuclear factor κB,1 and decreased COX-2 and endothelin. PPARγ ligands inhibit tumor growth and metastasis by inhibiting angiogenesis [116] because it is highly expressed in tumor endothelium in vivo and proliferating ECs in vitro. PPARγ ligands inhibit CAM angiogenesis and bFGF-induced corneal neovascularisation and prevent metastatic invasion. Another mechanism described in vitro experiments is to decrease metastasis via down-regulation of AKT [117]. The PPAR represents a large family of nuclear hormone receptors of 75 proteins in the mammalian proteome that enable the cells to respond to extracellular stimuli by transcriptional regulation of gene expressions [118]. Because these drugs have an excellent side-effect profile, the addition to a metronomic chemotherapy would therefore be easy to perform. Other approved drugs like nitroxolene [119] and rifampicin [120] have been described to feature anti-angiogenic properties. Nitroxolene is approved for the prophylaxis of bladder infection, thus continuous application should not be a problem. Rifampicin enters the enterohepatic circulation and undergoes hepatic accumulation, it may be beneficial as an anti-tumor agent targeting hepatobiliary tumors. Thalidomide, a sedative with anti-angiogenic properties [121], has been used in continuous low dosages in different cancer types with some therapeutic effects [122] but also with negative results [123]. The therapeutic effects of thalidomide in multiple myeloma as well as in myelodysplastic syndromes are possibly related to the anti-angiogenic properties of this drug. Similarly, long-term treatment with daily low doses of interferon α resulted in regressions of life-threatening hemangiomas [124]. It would be an interesting task to evaluate these non-cytotoxic agents which are approved drugs (for non-cancer indications) in combination with metronomic chemotherapy as well as in combination with `targeted drugs` because all these drugs can be orally administered in an essentially `metronomic` way.

Conclusion
Recent advances in anti-cancer treatment have been considerable for selected solid tumors due to new anti-doi: 10.7243/2049-7962-1-32 cancer drugs especially the introduction of so called `targeted drugs` such as monoclonal antibodies and small molecule signal transduction inhibitors (-nibs), however it is noteworthy to see that we still have ample room for improving our existing therapeutic armentarium to use 'old' approved drugs in new combinations, new schedules and new dosages. The focus in the past had been to obtain cures by tumor eradication at the expense of treatment-related complications/toxicity. However, the development of metronomic anti-cancer therapies is an approach designed to maintain a stable disease situation for advanced cancer patients (patients that cannot be cured) [125]. A metronomic CHT with a single drug is highly unlikely to be of benefit to the patients and to replace the conventional MTD defined CHT protocols when rapid tumor cell killing is necessary. The fundamental truth of oncology may still apply for the different metronomic concepts described in this paper: large tumors continue to be bad tumors and rapidly proliferating tumors and a high mutation rate will limit the treatment options a medical oncologist has, either using conventional MTD driven CHT or targeted therapy or metronomic therapy. Optimizing a metronomic anticancer therapy (which drugs ?, which tumors ? and when at which time point ?) is still an open question even after a decade of clinical work. Although the background of metronomic therapy is highly interesting, there is still a lack of good to excellent clinical trials. Phase III studies have not been performed and even high quality phase II studies are rare. This is actually the biggest hurdle for this low-cost and low-toxicity concept. It has to be shown that metronomic therapy is not a disappointing low-efficacy concept which only works in few cases as a palliative treatment. It is a great challenge to show that a mixture of drugs given in a metronomic way is effective and realistic step forward in controlling cancer.