Cancer Chemotherapy and Pharmacology

, Volume 72, Issue 1, pp 13–33

Metronomic chemotherapy for cancer treatment: a decade of clinical studies

Authors

    • Department of Oncology, Faculty of Medicine and PsychologySapienza University, Sant’Andrea Hospital
  • M. Christina Cox
    • Department of Hematology, Faculty of Medicine and PsychologySapienza University, Sant’Andrea Hospital
  • Ida Sarcina
    • Department of Oncology, Faculty of Medicine and PsychologySapienza University, Sant’Andrea Hospital
  • Roberta Di Rocco
    • Department of Oncology, Faculty of Medicine and PsychologySapienza University, Sant’Andrea Hospital
  • Chiara D’Antonio
    • Department of Oncology, Faculty of Medicine and PsychologySapienza University, Sant’Andrea Hospital
  • Viola Barucca
    • Department of Oncology, Faculty of Medicine and PsychologySapienza University, Sant’Andrea Hospital
  • Paolo Marchetti
    • Department of Oncology, Faculty of Medicine and PsychologySapienza University, Sant’Andrea Hospital
Review Article

DOI: 10.1007/s00280-013-2125-x

Cite this article as:
Romiti, A., Cox, M.C., Sarcina, I. et al. Cancer Chemother Pharmacol (2013) 72: 13. doi:10.1007/s00280-013-2125-x

Abstract

Purpose

Over the past few years, more and more new selective molecules directed against specific cellular targets have become available for cancer therapy, leading to impressive improvements. In this evolving scenario, a new way of delivering older cytotoxic drugs has also been developing. Many studies demonstrated that several cytotoxic drugs have antiangiogenic properties if administered frequently and at lower doses compared with standard schedules containing maximal tolerated doses (MTD). Such a new strategy, named metronomic chemotherapy, focuses on a different target: the slowly proliferating tumour endothelial cells. About 10 years ago, metronomic chemotherapy was firstly enunciated and hereafter many clinical experiences were published related to almost any cancer disease. This review analyses available studies dealing with metronomic chemotherapy and its combination with several targeted agents in solid tumours.

Methods

A computerized literature search of MEDLINE was performed using the following search terms: metronomic OR “continuous low dose” AND chemotherapy AND cancer OR solid tumours.

Results

Satisfactory results have been achieved in diverse tumour types, such as breast and prostate cancer or paediatric sarcomas. Moreover, many studies have reported that metronomic chemotherapy determined minimal toxicity compared to MTD chemotherapy. Overall, published series on metronomic schedules are very heterogeneous often reporting on retrospective data, while only very few studies were randomized trials. These limitations still prevent to draw definitive conclusions in diverse tumour types.

Conclusions

Large well-designed studies are eagerly awaited for confirming the promises of metronomic schedules and their combinations with targeted molecules.

Keywords

Metronomic chemotherapyCancer therapy“Continuous low dose” chemotherapySolid tumours

Introduction

Cancer therapy has impressively progressed over the past few years, with the development of targeted treatments [1]. In this new scenario, traditional cytotoxic chemotherapy has also been evolving. Both newer drug combinations and modalities of administration have been purposely tested to improve efficacy and tolerability of chemotherapies.

Conventional chemotherapy administration is generally based on the concept of the maximum tolerated dose (MTD), consisting in the highest survivable (minimum lethal) dose, steered to kill as many tumour cells as possible. However, this strategy, which was supported by very high cure rate in pre-clinical studies [2], requires treatment-free intervals to resume the growth of normal host cells, like the hematopoietic progenitors and the epithelial cells. Nonetheless, excluding some haematological or germinal malignancies that might be cured by conventional chemotherapy, the majority of cancer patients, after a period of disease regression or stabilization, experience tumour progression. The efficacy of conventional schedules is circumvented by several factors: the heterogeneity and the genomic instability of tumour cells, the protective action exerted by the microenvironment and the suppression of anticancer immune responses [3]. Moreover, tumour-associated angiogenesis, which is a pivotal step in tumour progression, is not efficiently affected by MTD chemotherapy. Endothelial cells in fact proliferate at a lower rate than tumour cells (chronic angiogenesis), this may account for the scarce activity of standard cytotoxic protocols on this compartment. Appealingly, it has been hypothesised that the recovery of endothelial cells occurring during treatment-free intervals supports the re-growth of tumour cells [4, 5].

At the end of the nineties, a few pivotal experimental studies showed that low doses of cytotoxic drugs, given with shorter intervals within consecutive doses and without interruption, exert a sustained cytotoxic or apoptotic effect on the tumour vascular endothelial cells, leading to tumour regression [4, 6]. As a result, a new logic of use for old drugs was elicited. This peculiar chemotherapy schedule, which directly or indirectly [5] also affects the slowly proliferating tumour endothelial cells, has been defined “Metronomic Chemotherapy” (MCT) [7]. After the era of the dose-intensity approach to cancer shrinkage, Kamen et al. [8] emphasize the emergence of a newer paradigm: high-time chemotherapy. High-time chemotherapy would seek the longest time for drug exposure at a given desired drug concentration to efficiently hit cancer cells. The Authors suggest reconsidering old drugs, as potential high-time therapy, to be used alongside of current strategies based on genomics, proteomics and informatics.

Actually, the awareness that drugs schedule was a critical determinant of therapeutic success preceded any notion that the target of “low dose”/“high-time” chemotherapy was the tumour vasculature. Several studies had previously reported encouraging clinical results on the chemotherapy delivered, following unconventional modalities such as oral daily dose, or continuous infusion (CI) and weekly schedules. These alternative schedules often proved effective and well tolerated [916].

Oversimplifying, it is retained that MTD regimens act on tumour cells, while metronomic or “high-time” therapies exert most of their effects on endothelial cells. An interesting work coupling MTD and MCT with the concurrent administration of multitarget kinase inhibitors in an animal model was the proof of this concept [17]. This combination resulted in an increased response rate and improved survival leading to a new chemo-switch model.

Metronomic chemotherapy: more than an antiangiogenic treatment

The search for more effective cancer treatments lead to a paradigm shift: cancer cells should not be dissociated from non-cancer cells present in the tumour microenvironment. Tumour outbreak and progression is in fact supported by complex interactions between tumour cells and non-malignant microenvironment. The fight against cancer might require to act against multiple possible targets: (1) cancer cells and cancer stem cells; (2) tumour vasculature; (3) tumour stroma, consisting in fibroblasts and specialized mesenchymal cells; (4) immune cells infiltrating the tumour; (5) the extracellular matrix.

Indeed, “metronomic scheduling of anticancer treatment” [18] allowing the continuous and simultaneous administration of several agents with different anticancer activities is currently looked at as a multitargeted treatment. The potentialities of this approach, also giving the broad spectrum of compounds that could be combined, yet need to be fully experimented and understood. Presently, several anticancer mechanisms are attributed to MCT schedule. Some of these activities have solid pre-clinical roots, while others, although theoretically sound, are still awaiting experimental evidence.

Antiangiogenic activity

The growth of some types of tumours may be subordinated both to angiogenesis (i.e. mature endothelial cell-dependent generation of new blood vessels) and to vasculogenesis (i.e. progenitor cell-dependent generation of new blood vessels). Compounds belonging to several classes of cytotoxic drugs have been reported to have one or more antiangiogenic effects [19]. Studies by both Browder et al. [4] and Klement et al. [6] stylishly demonstrated that, in mice xenografts of tumour cell lines, certain drugs express remarkable antiangiogenic effects if given at defined doses and schedules. The antiangiogenic activity was effective in tapering tumour growth and could overcome drug resistance [4]. The combination therapy of low-dose chemotherapy with antiangiogenic molecules such as TNP-470 [4] or DC101 [6] determined impressive results with full and sustained regressions of large established tumours. These experiments demonstrated that MCT exerts a sustained cytotoxic or apoptotic effect on the tumour vascular endothelial cells, leading to tumour regression. More recently, it was also reported that low metronomic doses of doxorubicin blocked HIF-1α binding to DNA and inhibited the transcription of correlated genes such as VEGF [20]. The inhibition of HIF-1α activity increased destruction of the tumour vasculature and overcame resistance to antiangiogenic therapies [21].

Restoration of cancer immunity

Several pre-clinical and clinical studies have shown that MCT administration of several cytotoxic drugs causes the selective depletion of immunosuppressive Treg cells. Tregs depletion allows the restoration of specific antitumour immunity by CD8 + -T cells, CD4 + -T cells and by tumour unspecified effector cells [22]. Other non-cytotoxic drugs like lenalidomide, thalidomide, metformine, COX-inhibitors may as well promote the activity of immune effectors. Another mechanism by which several cytotoxic drugs delivered with a MCT schedule might activate antitumour response is inducing the maturation of tumour-infiltrating dendritic cells [18].

Induction of tumour dormancy or quiescence

Tumour quiescence is considered a crucial mechanism in cancer patients, explaining both achievement of sustained remission and late relapse. Tumour dormancy is the result of cell cycle arrest or of a balance between proliferation and apoptosis. Three different mechanisms may contribute to this “steady-state:” angiogenic dormancy, immune surveillance and cellular dormancy. Although MCT has proven antiangiogenic and immunostimulating activities, presently there is no direct experimental evidence that MCT can induce tumour quiescence [22].

Direct effect on cancer cells

It has been theorized a “Drug-driven dependency/deprivation effect or 4D effect” of MCT [18]. This theory stems from combining the Darwinian rules to which cancer clones are exposed during cytotoxic treatment to experimental and clinical evidences. In fact, it was shown in vitro that drug-free intervals after long-term exposition may overcome or even revert drug resistance [23]. While in vivo, this theory is supported by the observation that several MCT protocols, that have proved very effective, foresaw the cyclic interruption of every single drug related to the MCT schedules [24, 25].

Reducing tumour stem-like cells (TSLC)

Self-renewal and differentiating capacities along with a slow proliferating rate, active DNA damage repair and antiapoptotic pathways, as also the expression of multidrug transporters are features associated with normal stem cells. Numerous studies indicate that the growth of gliomas and a number of other types of cancer are initiated and driven by a subpopulation of cancer cells with stem-like characteristics. Experimental evidences [26, 27] suggest that the vascular microenvironment has a role in maintaining brain TSLC identity and functions. The retention of such properties by TSLC also suggests that, like normal stem cells, TSLC may be inherently resistant to many traditional anticancer therapies that target rapidly dividing cells. Consistent with this notion, many studies show that proangiogenic growth factors such as basic fibroblast growth factor, epidermal growth factor and platelet-derived growth factor, which would ostensibly be in higher concentration near blood vessels, permit maintenance and expansion of glioma TSLC in culture. It was shown that factors secreted by endothelial cells increase the TSLC fraction and that the inhibition of angiogenesis by MCT plays also a key role in the elimination of TSLC [28].

It has already been proved in pre-clinical studies that MCT exerts a synergistic activity when combined with agents targeting neoangiogenesis or the immune system. Moreover, MCT schedules, giving their low toxic profile, are particularly suitable to be combined with other biological agents and also with conventional chemotherapy. These new possibilities widen the therapeutic scenario and call for further biological and clinical studies.

One decade elapsed from the intriguing enunciation of the metronomic concept and since then, several clinical studies have been accomplished with heterogeneous results. Existing reports on MCT and on MCT combined with targeted therapies in different types of solid tumours have been here reviewed with the purpose to disclose evidences and open questions.

Materials and methods

Search strategy, study selection criteria and data extraction

A computerized literature search of MEDLINE was performed using the following search terms: metronomic OR “continuous low dose” AND chemotherapy AND cancer OR solid tumours. English-written literature was considered valuable for the review. To identify additional studies, the bibliographies of the identified papers were searched for further relevant articles. From the review process were excluded all those studies in which it was not possible to extrapolate data on homogeneous diagnostic sub-groups, or detailed information on treatment protocols. Also, data reported only in abstract form were excluded.

Metronomic protocols in solid tumours

Central nervous system cancers

Adult patients

Glioblastomas are highly vascularized tumours with elevated expression of vascular endothelial growth factor (VEGF), thus providing a rationale for the use of antiangiogenic agents. At present time, the main treatment for glioblastoma multiforme (GBM) is still represented by surgery followed by radio-chemotherapy. Unfortunately, this aggressive treatment often allows only brief remissions with disappointing median progression-free and overall survival (6.9 and 14.6 months, respectively) [29]. For recurrent gliomas, median progression-free survival (PFS) and overall survival (OS) of 10 and 30 weeks were, respectively, reported [30]. Salvage therapy consists in re-challenging with temozolomide (TMZ) often incorporated in dose-intensified chemo-radiotherapy protocols. Several experiences aimed to elicit a synergic effect by combining antiangiogenic molecules with chemotherapeutic agents [31, 32]. After a pioneering study using daily low dose of Etoposide, which had yielded encouraging results in patients with malignant glioma [33], several studies followed. Those trials combined metronomic schedules of etoposide, temozolomide (TMZ) and cyclophosphamide (CTX) but globally taken, these early clinical experiences with metronomic schemes were disappointing (Table 1).
Table 1

Phase II studies concerning the use of chemotherapy with a metronomic schedule in recurrent central nervous system cancers

Metronomic protocol

Histology

Patients (N)

ORR (%)

Clinical benefit (%)

Median PFS (weeks)

6-month PFS (%)

Median OS (weeks)

1-year OS (%)

Reference

mCTX-mMTX

GBM

10

0

 

10*

0

28

 

Herrlinger [34]

mVP16-CTX-CXB-Thalidomide

GBM

28

2

61

11

9

21

 

Kesari [35]

AG

20

14

26

41

 

mVP16-Bev

GBM-AG

59

24

96

24

41

63

 

Reardon [36]

mTMZ

GBM

38

5

60

17

32

41

 

Kong [37]

mTMZ

AG

28

15

38

36

 

61

Perry [38]

GBM early

33

3

27

14

27

 

27

GBM extended

27

0

8

7

7

 

15

GBM rechallenge

28

11

37

15

36

 

29

mTMZ-Rofecoxib

GBM

13

8

 

32*

 

64

 

Tuettenberg [32]

mTMZ-CXB

GBM

28

 

54

17

43

17

 

Stockammer [39]

mVP16-Bev

GBM

13

0

62

8

8

19

 

Reardon [41]

mTMZ-Bev

GBM

10

0

40

4

0

13

 

mTMZ-Bev

GBM

32

28

 

16

19

37

31

Desjardins [42]

ORR overall response rate (complete response + partial response), clinical benefit (complete response + partial response + stable disease), PFS progression-free survival, OS overall survival, mCTX metronomic Ciclophosphamide, mMTX metronomic Metotrexate, mVP16 metronomic Etoposide, CXB celecoxib, mTMZ metronomic Temozolomide, Bev bevacizumab, GBM glioblastoma multiforme, AG anaplastic glioma

* Median time to progression

In details, patients with GBM or Anaplastic Gliomas (AG) were not responsive to both schemes using low doses of CTX combined with methotrexate (MTX) [34] and to a more complex antiangiogenic regimen consisting of low doses of etoposide and CTX plus daily thalidomide and celecoxib (CXB) [35]. Unsatisfactory results were also reported in patients with recurrent Glioblastoma when metronomic etoposide was combined to bevacizumab [36]. Actually, comparing the effects of this combination with historical data derived from patients treated with Bevacizumab alone, a similar efficacy but a greater toxicity could be noticed.

Recently, it was devised a protocol with TMZ administered with a protracted low-dose schedule [32, 3739] resulting in encouraging clinical benefit (CB: objective response plus stable disease) for these very poor prognosis patients.

In the RESCUE trial [38], patients with malignant glioma, who progressed after standard TMZ, were stratified by tumour type. GBM patients were divided into three groups, according to the timing of progression during adjuvant therapy. The authors found that the re-challenge with continuous daily TMZ was a valid therapeutic option for patients with recurrent GBM. Patients who experienced a disease progression during the first six cycles of conventional adjuvant TMZ or after a treatment-free interval had the major benefit from therapy. The results were comparable to those obtained with antiangiogenic therapies such as bevacizumab [40]; moreover, the tolerability was good, with minimal haematological and gastrointestinal toxicity.

However, two other phase II studies, combining MCT with bevacizumab, yielded disappointing results and did not reveal synergistic effects of the association [41, 42]. In the former, metronomic etoposide or TMZ was administered along with Bevacizumab in heavily pre-treated patients with recurrent GBM, who had also progressed after a scheme containing bevacizumab, with poor results [41]. Both etoposide and TMZ arms of the study closed at the interim analysis, due to the lack of adequate antitumour activity. In the latter study [42], PFS and OS in GBM patients, previously treated with radiotherapy and TMZ, did not compare favourably with those reported for bevacizumab monotherapy or bevacizumab plus irinotecan [43]. Even though some high-grade toxicity events were reported, overall, such treatments were considered well tolerated.

Finally, a randomized study investigated, in adult patients with newly diagnosed GBM, two TMZ regimens, as adjuvant treatment after standard radiotherapy with concurrent daily TMZ [44]. The continuous daily TMZ determined a worse 1-year survival rate compared to the dose-dense TMZ schedule (69 vs. 80 %). Both dose-dense and metronomic TMZ regimens were well tolerated with modest toxicity.

Globally taken, metronomic schedules using single drugs or combinations with bevacizumab did not yield persuasive results. Nonetheless, given the low toxic profile and the activity of some metronomic schedules [3739], we believe that further protocols including metronomic TMZ deserve to be explored, possibly in randomized studies.

Paediatric patients

Children less than 5 years of age with CNS tumours have a high rate of morbidity and mortality following the administration of conventional therapy. Older children who have recurrent or refractory CNS cancer also fare poorly with conventional treatments.

MCT schedules have been investigated in these settings since 2006 when Sterba et al. [45] set-up a four-drug protocol named COMBAT, which is an acronym for “combined oral maintenance bio-differentiating and antiangiogenic therapy.” Treatment consisted of daily CXB administration along with daily 13-cisretinoic acid and cycles of metronomic TMZ and low-dose etoposide. Each drug had treatment-free intervals and the schedule was devised for a period of one year. Overall, nine out of 14 patients assessable for response had prolonged CB, and these responding patients were children (4- to 11 years old) with a progressive CNS tumour. In the same study, another group of older children (10–17 years) with medulloblastoma in CR after more than one line of therapy were also treated with the COMBAT protocol as a maintenance treatment. The overall toxicity was mild and the treatment well tolerated requiring slight reductions. Worthy of note in several patients response duration was longer than previously achieved with the administration of MTD treatments.

In 2008, the retrospective data of ten children less than 5-years-old with metastatic central nervous system tumours, who were given MCT schedule in the setting of minimal residual disease after high-dose chemotherapy (HDCT), were reported [24]. Treatment was again a complex combination alternating three different cytotoxic drugs with two different biological agents. The event-free survival (EFS) was very satisfying compared to historical controls. Overall, except one case, none of the patients had significant infections, hospitalizations or haematological toxicity during MCT administration. In all patients, an excellent quality of life was maintained throughout. This interesting preliminary study suggests potential benefit for MCT chemotherapy as a remission maintenance strategy following first-line HDCT. The third paper is a phase II prospective study: 31 children (median age 10.5 years) with refractory or recurrent CNS tumours were treated with continuous low dose of oral topotecan [46]. Twenty-five patients were evaluable for response: 20 % had stable disease (SD) at first evaluation and 8 % had partial response (PR). In four out of five patients in SD, the steady state lasted 3–7.6 months on therapy. The two patients who achieved partial remission had disseminated medulloblastoma at study entry. These two patients were on treatment for 6 and 15 months respectively. They continue to have clinical and radiographic remission lasting more than 7 and 10 years, respectively, from discontinuation of oral topotecan without further therapy. The remaining 18 patients had disease progression within three months from the start of treatment. Very recently were published the results of a trial combining multiple cytotoxic molecules and biological agents in children with recurrent embryonal brain tumour (median age 9 years) [25]. An antiangiogenic drugs combination containing CTX, etoposide, bevacizumab, thalidomide, CXB, fenofibrate and dexamethasone was administered. In addition, etoposide and liposomal cytarabine were given as intraventricular therapy with an Ohmamaya reservoir. Very good results were obtained in the seven patients with medulloblastoma: at 24 months, both OS and EFS were 68.6 ± 19 %. In contrast for patients with PNET, results were disappointing: OS and EFS after 6 months were 75.0 ± 22 % and 0.0 %, respectively, and all patients had died by 12 months. Presently, a formal phase II study [NCT01356290] is ongoing in medulloblastoma patients.

Breast cancer

Breast cancer incidence is stable or increasing in many western countries. Over the last decades, it was however registered an encouraging rise in survival rates. This is partially the result of a complex interplay among different therapeutic disciplines. Surgery, radiotherapy, hormonal therapies and chemotherapy along with the newer targeted therapies all contribute to the cure of breast cancer patients. Nonetheless, when the disease recurs after primary treatments, in spite of the use of the most effective chemotherapies, the PFS ranges between 5 and 12 months [4749]. Moreover, PFS gets lower when chemotherapy is used in patients heavily pre-treated for advanced disease [50, 51].

Many drugs currently used in standard protocols as CTX [5263], MTX [5359], fluoropyrimidines [6466] and VNR [67] have been also experimented in metronomic schedules in different clinical setting and histological subtypes. These metronomic protocols were sometimes associated with hormones [68, 69] or with targeted agents like trastuzumab [59] and bevacizumab [7073] or vaccines [74] (Table 2).
Table 2

Clinical studies concerning the use of chemotherapy with a metronomic schedule in breast cancer

Metronomic protocol

Study design

Patients (N)

ORR (%)

Clinical benefit (%)

Median TTP (months)

2-year-DFS (%)

Median OS (months)

1-year OS (%)

Reference

mCTX

Observational

12

0

58

  

14

 

Ge [52]

mCTX + mMTX

Phase II

64

19

32

3

  

26

Colleoni [53]

mCTX + mMTX

Retrospective

62

  

3

 

7

 

Miscoria [54]

mCTX

Retrospective

22

14

55

4

 

13

 

Gebbia [55]

mCTX-mMTX

39

20

51

4

 

14

 

mCTX-mMTX versus

Phase III

86

21

41

4

 

18

 

Colleoni [56]

mCTX-mMTX- Thalidomide

85

12

41

4

 

17

 

mCTX-mMTX-Dalteparin-Prednisone

Phase I/II

41

17

24

2

 

12

 

Wong [57]

mCTX-mMTX-CXB

Phase II

67

0

31

2

 

8

 

Khan [58]

mCTX-mMTX-Trastuzumab

Phase II

22

18

46

6

   

Orlando [59]

mCTX-EPI-5FU-VCR-Prednisone

Retrospective

63

59

94

14

 

28

 

Gonzales-Billalabeitia [60]

mCTX-Pegylated Doxorubicin

Phase II

29

62

96

    

Dellapasqua [61]

mCTX-CAP

Phase II

66

30

53

5

 

17

 

Wang [62]

mCTX-CAP

Phase II

45

44

58

12*

 

Not reached

 

Yoshimoto [63]

mCAP

Phase II

60

24

62

7

 

17

 

Fedele [64]

m5FU-Eniluracil

Phase II

33

48

78

7

   

Smith [65]

mCAP-mTXT-CXB

Phase II

38

34

42

4

 

10

 

Young [66]

mVNR

Phase II

34 elderly

38

70

8

 

16

 

Addeo [67]

mCTX-Megestrol acetate

Phase II

29

31

41

7

 

13

 

Licchetta [68]

mCTX-Letrozole versus Letrozole

Phase II R

57

88

97

 

82

  

Bottini [69]

57

72

95

 

83

  

mCAP-mCTX-Bev

Phase II

46

48

68

10

   

Dellapasqua [70]

mCAP-mCTX-Bev-Erlotinib

Phase II

24

62

75

11

 

25

 

Montagna [71]

mCTX-mMTX-Bev

Phase II

24

32

64

7*

 

14

 

Garcia-Saenz [72]

mVNR-Bev

Phase II

13

8

54

4*

  

74

Saloustros [73]

ORR overall response rate (complete response + partial response); clinical benefit (complete response + partial response + stable disease), TTP time to progression, DFS disease-free survival, OS overall survival, mCTX metronomic Cyclophosphamide, mMTX metronomic Metotrexate, CXB celecoxib, m5FU metronomic 5-Fluorouracil, EPI Epirubicin, VCR Vincristin, mVNR metronomic Vinorelbin, mCAP metronomic Capecitabine, Bev bevacizumab

* Median progression-free survival

Several studies investigated the association of metronomic CTX and MTX. The reported overall response rate (ORR), CB and median OS ranged, respectively, between 19–21, 24–51 % and 12–18 months [5355]. The adjunct of antiangiogenic-immunomodulating, anti-inflammatory agents or trastuzumab to this schedule did not allow substantial improvement of results [5659].

Metronomic CTX was also combined with standard-dose anthracyclines [60, 61]. The reported ORR was 59 % in the neo-adjuvant setting (61) and 62 % in elderly or unfit patients with locally advanced breast cancer (60). Moreover, two recent studies tested metronomic CTX in adjunct to capecitabine [62, 63]. Tolerability was generally good and a satisfactory percentage of CB was achieved both in pre-treated patients irrespective to HER2 status [62] and in HER2-negative metastatic breast cancer patients [63].

Oral fluoropyrimidines have also proved effective in metronomic protocols in heavily pre-treated MBC patients, with a reported PFS up to 7 months [64, 65]. A metronomic schedule of both docetaxel and capecitabine plus CXB was tested in anthracycline-resistant or -refractory, locally advanced or metastatic disease, to confirm their potential synergism, which was suggested by pre-clinical observations [66]. However, efficacy fell short of expectations, with CB below 50 %, that was the primary end point of the study.

On the other hand, metronomic oral vinorelbine (VNB) showed both a promising activity and a good tolerability in elderly patients with MBC [67].

The metronomic CTX was also coupled with hormones. The combination of oral daily CTX with daily, fractionated doses of megestrol acetate was well tolerated and determined CB in 41 % of advanced BC patients [68]. A randomized study, investigating the adjunct of metronomic CTX to letrozole as a neo-adjuvant protocol in elderly patients, showed a statistically significant advantage in ORR for the association schedule compared with letrozole alone [69]. Moreover, the evaluation of VEGF-A on tumour tissue was significantly lower in letrozole-CTX patients than in Letrozole ones, confirming the antiangiogenic property of mCTX.

Other attempts focused on achieving CB in MBC patients heavily pre-treated or refractory to anthracycline or taxanes, by coupling MCT and bevacizumab [7073].

Worthy of note, in a study enrolling pre-treated advanced or MBC (76 % hormone receptor positive), it was explored the activity of a new regimen including metronomic capecitabine and CTX plus bevacizumab [70]. Moreover, circulating endothelial cells (CECs) and circulating endothelial progenitors (CEPs) were investigated as surrogate biomarkers of response. A clinically relevant fraction of the patients (68 %) achieved a control of the disease for at least 6 months without significant acute or delayed toxicity. The study also indicated that high baseline CECs levels significantly correlated with the achievement of clinical response (P = 0.02) and CB (P = 0.01), therefore supporting their predictive role for antiangiogenic treatments.

Recently, these authors also tested, in a selected group of MBC with HER-2-negative MBC and poor hormone receptor expression, the same regimen plus erlotinib as first-line therapy [71]. They would explore the hypothesis that an increase in antitumour activity might be achieved by targeting both the EGFR and VEGF pathways. Truly, the study reported an absolutely good ORR (62 %), demonstrating that a multitarget protocol including MCT may be an effective therapeutic option not only for patients with indolent and low-proliferating tumours, such as the hormone receptor positive ones, but also for those triple negative or with poor hormone receptor expression. Moreover, the study also showed that low baseline CEPs levels were predictive of clinical responses.

Some anecdotal experiences have been recently published using metronomic CTX and immunotherapy but their results need to be more extensively validated [74].

Finally, a recent phase III trial showed oral uracil-tegafur (UFT) was as effective as classical CMF as adjuvant treatment, in women with node-negative, high-risk BC [75]. The QOL scores were better for patients given UFT than those given CMF. These encouraging results were confirmed in a pooled analysis demonstrating that the oral UFT is non-inferior to CMF (6 courses) in terms of inhibiting recurrence of ER-positive, early BC [76].

Globally taken, these heterogeneous studies support the clinical efficacy of MCT in BC. This was shown even in the setting of advanced, metastatic or refractory disease. Notably, the combining of cytotoxic drugs at metronomic dosages with targeted agents yielded very promising results [70, 71]. However, more robust data are expected at the end of some ongoing phase III trials. The study NCT01131195 will provide indeed results about tolerability and efficacy of a combination of bevacizumab plus a metronomic schedule of CTX and capecitabine compared to bevacizumab plus weekly paclitaxel in MBC patients. The NCT01112826 will clarify instead the impact of MCT after a standard adjuvant chemotherapy on DFS in triple-negative breast cancer

Lung cancer

Lung cancer represents the main cause of cancer death in western countries. Surgery, when possible, provides a chance of cure. In more advanced stages, however, both chemotherapy and radiotherapy play the leading, therapeutic role. In local advanced and metastatic non-small cell lung cancer (NSCLC), best chemotherapy schedules, including platinum compounds and targeted therapy, led to a median PFS and OS up to 5 and 14 months, respectively [7779]. Better results, with ORR of 67 %, median time to progression (TTP) and OS of 12 and 24 months, respectively, were achieved using tyrosine kinase inhibitors (TKIs) in those histotypes harbouring EGFR mutations [80]. The outcomes of standard therapy sensibly worsen proceeding to successive treatment lines, with ORR less than 10 % and disappointing, median PFS and OS of about 2 and 6 months [81].

Apart a randomized controlled trial of adjuvant chemotherapy including oral, UFT continuously delivered over 2 years, which yielded positive results [82], only few data concern phase II studies of MCT in advanced NSCLC (Table 3). In a pilot study, a combination of weekly cisplatin and daily oral etoposide was tested in high-risk (stages IIIB–IV) patients with NSCLC [83]. The encouraging results led to a phase Ib/II trial focused on the same setting of patients [84, 85]. They all received a schedule combining standard doses of cisplatin plus bevacizumab and metronomic doses of oral etoposide every 3 weeks. Both responsive patients and patients with SD (overall 87 %) received maintenance treatment with bevacizumab in combination with erlotinib [85]. The regimen proved very active but poorly tolerated with an ORR of 69 % and several G3-4 toxicities. A pilot study tested a doublet of weekly docetaxel and metronomic trofosfamide, in pre-treated stage IV NSCLC patients [86]. In this case, the tolerability was good but the ORR was only 19 %. Another study investigated serum concentrations of some biomarkers of angiogenesis as VEGF, thrombospondin-1 (TSP1) and VEGFR1 in NSCLC patients treated with either MTD or low dose of cisplatin and docetaxel [87]. Disappointingly, the metronomic schedule failed to show more pronounced antiangiogenic effects than the same doublet administered at MTD.
Table 3

Clinical studies concerning the use of chemotherapy with a metronomic schedule in lung cancer, sarcoma and melanoma

Metronomic protocol

Study design

Histology

Patients (N)

ORR (%)

Clinical benefit (%)

Median PFS (months)

Median OS (months)

Reference

CDDP-mVP16

Pilot

NSCLC

31

45

58

9

13

Correale [83]

CDDP-mVP16-Bev

Phase II

NSCLC

45

69

87

9

12

Correale [85]

Docetaxel-mTrofosfamide

Pilot

NSCLC

21

19

 

3

7

Gorn [86]

mTMZ

Phase II

NSCLC

31

6

16

2*

3

Kouroussis [89]

Gefitinib

Phase II R

NSCLC

58

35

66

5

18

Chen [90]

Gefitinib-mUFT

57

37

76

8

24

mTrofosfamide

Phase II

STS

18

 

50

3

7

Hartmann [94]

mTrofosfamide-

Case series

AS

5

50

67**

8

 

Vogt [95]

Pioglitazone-Rofecoxib

EHE

1

 

mTrofosfamide-Pioglitazone-Rofecoxib

Phase II

Sarcoma

21

19

33

4

6

Reichle [96]

mVP16

Retrospective

STS

26

4

46

5

10

Italiano [98]

mCTX-Prednisolone

Retrospective

STS

26 elderly

27

69

7

14

Mir [99]

mCTX-

Retrospective

Melanoma

13 elderly

8

46

2*

8

Borne [105]

mTreosulfan-Rofecoxib

Pilot

Melanoma

12

8

42

9*

13

Spieth [106]

mTMZ-Thalidomide

Phase II

Melanoma

38

32

74

 

9

Hwu [107]

mTMZ-Thalidomide

Phase II

Melanoma

64

13

 

2

8

Clark [108]

mPTX-CXB

Pilot

Melanoma

20

5

20

2

7

Bhatt [109]

Trofosfamide

Phase II R

Melanoma

32

3

 

1

8

Reichle [110]

Trofosfamide-Pioglitazone-Rofecoxib

35

3

 

19

mCTX-mMTX-CXB

Phase II

Melanoma

15

0

27

2*

5

Khan [58]

ORR overall response rate (complete response + partial response), CB clinical benefit (complete response + partial response + stable disease), PFS disease-free survival, OS overall survival, mCDDP metronomic Cisplatin, mVP16 metronomic Etoposide, Bev Bevacizumab, mTMZ metronomic Temozolomide, mUFT metronomic Tegafur-Uracil, mCTX metronomic Ciclophosphamide, mPTX metronomic Paclitaxel, CXB Celecoxib, NSCLC non-small cell lung cancer, STS soft tissue sarcoma, AS angiosarcoma, EHE hemangioendotelioma

* Median time to progression; ** CB lasted for >6 months; P > 0.05

Metronomic VNR has been recently evaluated in association with standard cisplatin dosage in inoperable, locally advanced or metastatic NSCLC patients, who had received at least two prior chemotherapy lines [88]. This regimen, considering the set of patients included, showed a not negligible activity, with a 21 % of objective responses and a good tolerability profile. Another trial, using metronomic TMZ as salvage therapy in patients with advanced NSCLC, showed only a minimal clinical activity [89]. Finally, a phase II trial randomized 115 patients, who had failed previous treatments, to receive either gefitinib alone or in combination with metronomic UFT [90]. Worthy of note, the addition of UFT significantly improved PFS in those patients harbouring EGFR mutations (14.4 vs. 7.6 months, P < 0.01) while no survival advantage was in patients with both wild-type and unknown EGFR status.

The few studies dealing with MCT schedules in NSLC yielded some encouraging results with ORR up to 69 % [85]. Notably, effective protocols consisted of a combined schedule containing or standard-dose cisplatin [83, 85] or targeted therapies [90].

Sarcomas

Adult patients

Sarcomas are an heterogeneous group of rare diseases. Altogether, soft tissue sarcomas (STS) represent the prevalent subtypes in adults. Chemotherapy exerts a pivotal role in advanced, unresectable or metastatic disease. Nonetheless, in these settings, the ORR is less than 20 % and the median PFS and OS are only 6 and 18 months, respectively [91]. Relapsed patients treated with either cytotoxic agents [92] or newer targeted therapies [93] have an appalling outcome with a median OS of 8–12 months.

A phase II study suggested a role for trofosfamide as an effective palliative treatment for patients with STS [94] (Table 3). Eighteen heavily pre-treated patients with STS were given oral, continuously delivered trofosfamide. No objective remission was observed but almost half of the patients achieved SD for 6 months with an acceptable toxicity profile.

Metronomic trofosfamide plus pioglitazone, a peroxisome proliferator-activated receptor gamma agonist, and the cyclooxygenase-2 (COX-2) inhibitor rofecoxib, were investigated in two different studies enrolling patients with chemo-refractory STS [95, 96]. In the first study, two out of six patients with advanced vascular sarcomas achieved complete remission while toxicity was mild [95]. These positive results in such critical patients were also confirmed in the second study [96].

Other trials reported a moderate activity of metronomic vinorelbine [97] and etoposide [98] in refractory sarcomas. A recent study evaluated daily low dose of CTX and prednisolone in elderly patients with unresectable STS [99]. About 70 % of the patients achieved CB with moderate toxicity. Median PFS was significantly longer in those patients with radiation-induced sarcomas (7.8 vs 5.2 months).

Paediatric patients

In 2006, Sterba et al. [45] reported a series of children with various types of sarcomas who were treated with the COMBAT protocol. In this trial, two children with recurrent osteosarcoma and one with Ewing’s sarcoma were treated: only one had CB.

After a pilot study designed to define the optimal dose of intravenous vinorelbin combined with oral low-dose CTX in patients with recurrent or refractory STS [100], in 2012, Minar-Colin et al. [101] reported the results of a large phase II study adopting the same protocol. Of 117 patients enrolled with mixed sarcoma, 43 % were Rhabdomyosarcomas (RMS). The median age was 12, while age range was 1–24 years. All patients were relapsed or refractory after one (43 %), or more than one line of treatment (56 %). Patients with RMS (N = 50) had an impressive ORR of 36 %, including 14 % PR and 8 % CR. SD was observed in 24 %. Patients with untreated relapse were the best responders (ORR 45 %). The median survival time for the whole population with RMS was 9 months (95 % CI 6–12 months). In non-RMS STS and in all other diagnostic groups, responses were very rare. The overall tolerability profile was good, and grade 3/4 myelosuppression was the most common adverse event, occurring in 38 % of patients; nonetheless, granulocyte stimulating factors were not used.

Melanoma

Metastatic melanoma has a dismal prognosis. Conventional chemotherapy, including dacarbazin, TMZ, paclitaxel or platinum compounds, allows an ORR not reaching 20 %. Recently, very positive results have been achieved with new targeted agents. In a first study, ipilimumab, a monoclonal antibody directed to “cytotoxic T lymphocyte antigen-4,” combined with a glycoprotein 100 peptide vaccine (gp100) induced a survival advantage in metastatic melanoma compared to gp100 alone (OS: 10.1 vs 6.4 months, respectively, HR: 0.066, P = 0.003) [102]. In a second study, ipilimumab combined with dacarbazin induced a better OS compared to dacarbazin alone (11.2 vs 9.1 months) [103]. Moreover, vemurafenib, a specific inhibitor of signalling by mutated B-Raf, significantly improved OS (HR: 0.37, P < 0.001) along with PFS (HR:0.26, P < 0.001) in respect of the dacarbazin arm in another phase III study [104].

Data about MCT in melanoma patients are mostly derived from pilot studies including only few cases (Table 3). Metronomic CTX led to disease control in about 50 % of elderly, melanoma patients while RR and survival were comparable to those observed with conventional chemotherapy such as dacarbazine [105]. A similar disease control had been previously reported using another alkylating agent, treosulfan combined with rofecoxib [106].

Daily, low-dose TMZ plus concomitant thalidomide (dose adjusted according to age) demonstrated significant clinical activity in patients with advanced-stage or metastatic melanoma [107]. Thirty-two per cent of patients achieved an objective tumour response while treatment was generally well tolerated. Unfortunately, a more recent, multicentric, phase II trial, which used the same chemotherapy protocol, reported conflicting results [108]. The ORR was really low, while some G4 toxicities, including one treatment-related death, were reported. Several cytokines were also evaluated to assess the immune modulatory effect of thalidomide in combination with TMZ. However, none of these biomarkers was found to be related with survival parameters, at a significance level of 0.05.

Other studies experimented the adjunct of CXB to metronomic paclitaxel [109] or CTX plus MTX [58] to control tumour growth and angiogenesis in melanoma patients. These multitarget therapies however exhibited a disappointing lack of clinical activity with ORR below 10 %. The pilot study by Bhatt et al. [109], which enrolled patients with refractory metastatic melanoma, also evaluated a series of biomarkers, such as CEC/CEP, sVEGFR2, IL8, angiopoietin 2, fibroblast growth factor, VEGF and TSP1. Probably, due to the exiguity of the sample size, there was no correlation between each of the biomarkers levels and clinical outcomes.

In a randomized phase II trial, metronomic trophosphamide was compared to a combination of the same drug plus the anti-inflammatory molecules pioglitazone and rofecoxib [110]. Seventy-six patients with metastatic melanoma were randomized: the concurrent inhibition of tumour-associated inflammatory processes with anti-inflammatory molecules resulted in a longer PFS (HR 2.13 (95 % CI 1.29–3.51), P = 0.008). Interestingly, the authors demonstrated that the addition of a combined anti-inflammatory treatment more often prompted a reduction in C-reactive protein (CRP) levels in comparison with MCT alone, and that the CRP response was closely correlated to OS. Treatment was well tolerated in both subgroups of patients without G4 toxicities.

In conclusion, despite some positive results, the overall outlook of metronomic schedules is fairly disappointing: reported median PFS rarely exceeded 3 months in the largest series.

Gastrointestinal tumours

Most data concerning MCT in gastrointestinal cancer derive from hepatocellular carcinoma (HCC) and colorectal cancer (CRC) trials (Table 4).
Table 4

Clinical studies concerning the use of chemotherapy with a metronomic schedule in advanced gastrointestinal cancer

Metronomic protocol

Study design

Histology

Patients (N)

ORR (%)

Clinical benefit (%)

Median PFS (months)

Median OS (months)

Reference

mUFT-Sorafenib

Phase II

HCC

53

8

57

4

7

Hsu [112]

mUFT-THAL

Phase II

HCC

43

9

42

2

5

Shao [113]

OCT

Phase I/II R

HCC

10

  

4*

9

Treiber [114]

OCT-IM

11

  

3*

8

OXL (60–90 mg)

7

  

2*

16

OXL (20–30 mg)-OCT-IM

10

  

3*

12

mCAP-CXB

Phase II

mCRC

12

 

34

4*

 

Steinbild [118]

Pancreato-biliary

7

  

mCTX-Rofecoxib-mVBL

Phase II

GIC

23

13

30

3*

8

Young [119]

Other ST

27

    

mCTX-mMTX-CXB

Phase II

GIC

17

0

41

2

8

Khan [58]

mCTX-UFT-CXB

Phase II

GIC

38

 

45

3

7

Allegrini [120]

mCPT11-m Doxifluridine

Phase II

mCRC

45

36

73

6

15

Ogata [121]

OXL-FUFA → mUFT

Phase II

mCRC

28

36

 

5

13

Lin [122]

mCAP

Phase II

GC

45 elderly

21

51

4

8

He [127]

5FUc.i.-OCT

Phase II

NET

29

24

93

23*

Not reached

Brizzi [128]

ORR overall response rate (complete response + partial response); clinical benefit (complete response + partial response + stable disease), PFS progression-free survival, OS overall survival, mUFT metronomic Tegafur-Uracil, THAL Thalidomide, OCT Octreotide, IM Imatinib, OXL Oxaliplatin, CXB Celecoxib, mCAP metronomic Capecitabine, mCTX metronomic Ciclophosphamide, mVBL metronomic Vinblastin, mCPT11 metronomic Irinotecan, FUFA 5Fluorouracil + Folinic Acid, HCC hepatocellular carcinoma, mCRC metastatic Colorectal Cancer, GIC gastrointestinal cancer, ST, solid tumours, GC gastric cancer, NET pancreatic and gastrointestinal neuroendocrine tumours

* Median time to progression

HCC is a highly vascularised, poorly chemo-sensitive neoplasm. Theoretically, HCC might benefit from antiangiogenic drugs, but actually only sorafenib increased median OS in a randomized controlled trial (10.7 vs. 7.9 months in the placebo group; HR: 0.69; 95 %, P < 0.001) [111]. Sorafenib, delivered at 400 mg twice daily until progression, is presently standard of care for HCC.

A phase II study evaluated metronomic UFT combined with standard-dose sorafenib in patients with advanced HCC [112]. Previously, untreated patients with locally advanced or metastatic diseases, not amenable to loco-regional therapies, were enrolled. The ORR was slightly higher than previously reported for Sorafenib alone [111] while the toxicity profile was not increased.

In a multicenter phase II study, both safety and efficacy of adding metronomic UFT to thalidomide have been recently investigated in patients with advanced HCC non-amenable to loco-regional therapies [113]. Forty-two per cent of patients experienced CB but results concerning PFS and OS were dismal. In two subsequent prospective, randomised, phase I-II trials, the impact of different treatments on plasma biomarkers related to angiogenesis such as s-E-selectin, VEGF-A and PDGF-BB, was assessed [114]. The first trial compared octreotide alone with octreotide plus imatinib, being hypothesised that octreotide would lead to a decrease in VEGF-A, while imatinib would decrease PDGF-BB or attenuate its activation during chemotherapy. In the second trial, patients received oxaliplatin alone, at a standard dosage, or oxaliplatin, at a low weekly dose, combined with octreotide and imatinib. The hypothesis tested in this case was that MTD chemotherapy would increase all markers of angiogenesis compared to MCT. Indeed, the major increase in plasma levels of s-E-selectin and PDGF-BB were found in patients receiving MTD chemotherapy alone. Conversely, during MCT combined with antiangiogenic drugs, the biomarker levels had the minor increase. The adjunct of MCT to the targeted molecules, in an indirect comparison between the two studies, did not allow increased TTP and OS.

Standard chemotherapy of metastatic CRC (mCRC) consists of combinations of cytotoxic drugs such as fluoropyrimidines, oxaliplatin and irinotecan along with either anti-VEGF (bevacizumab) or anti-EGFR (cetuximab, panitumumab) monoclonal antibodies. These drugs raised the median OS over 24 months in selected patients [115]. However, patients with mCRC, already treated with both oxaliplatin and irinotecan, showed a dismal median OS (6 months) in spite of the use of new targeted therapies [116]. In recent years, a blooming of studies on the use of chemotherapy and targeted molecules has generated relevant clinical results in colorectal tumours; nevertheless, only few trials concerning metronomic schedules have been carried out.

A phase I study investigated both the pharmacokinetics and pharmacodynamics of metronomic irinotecan at different dose levels as i.v. infusion, along with changes in antiangiogenic (TSP-1) and proangiogenic factors (VEGF) in 20 CRC patients, heavily pre-treated with both irinotecan and oxaliplatin chemotherapy [117]. An increase in gene expression and plasma concentration of TSP-1 was found in patients receiving the lowest doses of irinotecan. This finding supports the in vivo antiangiogenic effect of irinotecan given at metronomic dose. The combination of continuous, low dose of capecitabine and CXB was investigated in a small casuistic including pre-treated patients with locally advanced or mCRC [118]. A disease stabilization was achieved in about 30 % of the cases, along with an antiangiogenic effect revealed by a dynamic contrast-enhanced magnetic resonance imaging. Another trial enrolled patients with advanced solid tumours, relapsed or refractory after at least one prior chemotherapy regimen, with poor clinical conditions [119]. The patients were treated with daily oral CTX and rofecoxib in addition to weekly vinblastine. Half of them also received oral minocycline with the intent of further inhibiting tumour angiogenesis. Unfortunately, the primary end point of CB was not reached, while severe haematological toxicity occurred in 25 % of the cases. Similar results but with a lower incidence of toxicities were reported in patients with gastrointestinal cancer with other multitarget regimens, including either metronomic CTX plus MTX and CXB [58] or metronomic CTX plus UFT ad CXB [120]. Particularly, the study by Allegrini et al. [120] not reporting any toxicities above grade I in patients with heavily pre-treated advanced malignancies, highlighted one of the main features of a metronomic protocol. Moreover, interesting results in terms both of activity and tolerability were obtained in other two studies [121, 122]. In the first one, patients with mCRC received weekly irinotecan and oral, daily doxifluridine, an intermediate metabolite of capecitabine [121]. In the second trial, pre-treated patients with mCRC received an infusional protocol with oxaliplatin, leucovorin and 5-FU followed by oral, daily UFT/leucovorin (122). Overall, grade 3 toxicity occurred in less than 5 % of the patients.

Because the optimal duration of systemic treatment in patients with advanced colorectal cancer is still unknown, a randomized phase III trial was planned to investigate the role of bevacizumab and low-dose continuous capecitabine, as maintenance chemotherapy, in patients with MCR, who have responded to an induction treatment at standard doses. The study (NCT00442637) is ongoing but not still recruiting participants.

Finally, an interesting, prospective study was conducted to evaluate the feasibility and the antitumour activity of MCT in post-operative setting [123]. Patients with CRC at high risk of recurrence received weekly CPT-11 plus oral, daily UFT, after a radical tumour resection. Cycles were repeated for 6 months and followed by UFT alone for further 6 months. Really, remarkable 5-year OS rates of 73 and 62 % were observed, respectively, in the stage IIIB and in the distant metastases groups. Toxicities were mild: G3 nausea in 5 % and none G4 toxicity.

Chemotherapy can provide palliation and improved survival compared to best supportive care in patients with advanced and metastatic gastric cancer (GC) [124]. Despite promising results in pre-clinical studies have been reported [125, 126], only one clinical study has evaluated MCT in GC [127]. In this experience, elderly patients with advanced GC, pre-treated with a fluoropyrimidine-based regimen, received metronomic capecitabine. About a half of these frail patients received a CB at 8 weeks while none was affected by Grade 4 toxicity and less than 10 % had a Grade 3 toxicity.

Finally, a multicenter phase II trial tested a continuous 5-Fluorouracil infusion plus LAR octreotide as upfront treatment for advanced, well-differentiated, pancreatic or gastrointestinal neuroendocrine carcinomas [128]. The rationale of the study is based on the given, high vascularization of well-differentiated neuroendocrine carcinomas. The regimen proved to be active and well tolerated. In fact, ORR was higher than that historically reported with somatostatin alone, suggesting a possible role of the continuous infusion of 5-FU associated with somatostatin in controlling tumour progression. Moreover, a stepwise reduction in the plasma VEGF levels was observed, reflecting the antiangiogenic activity of the protocol.

Prostate cancer

Prostate cancer is the most frequent male cancer in many western countries. The androgen deprivation therapy is the first-choice treatment for advanced and metastatic disease, while chemotherapy is used to treat symptomatic patients with castration-resistant prostate cancer. As first-line treatment, docetaxel produced a 2-month survival advantage relative to mitoxantrone, in two randomized trials [129, 130]. In patients with metastatic, castration-resistant prostate cancer, who previously received docetaxel, newer molecules, such as carbaxitaxel and abiraterone acetate allowed a statistically significant survival advantage, respectively, compared with mitoxantrone or placebo [131, 132]. Both PFS (5.6 vs. 3.6 months; P < 0.001) and PSA response rate (29 vs. 6 %, P < 0.001) were even improved by abiraterone compared with placebo [132]. The drop of PSA and the reduction of disease-related symptoms, like bone pains, are commonly used to assess treatment outcome in prostate cancer patients. Several studies also exploited these markers to demonstrate MCT activity. Oral metronomic CTX proved safe, well tolerated and effective in hormone-refractory advanced prostate cancer (HRPC) [133] (Table 5). Likewise, a group of studies analysing the association of metronomic CTX and dexamethasone or prednisolone in HRPC reported a PSA decrease as high as 50 % in up to 69 % of the patients [134137]. In a study, evaluating in HRPC patients, a schedule with a single i.v. standard dose of CTX immediately followed by metronomic CTX plus CXB and dexamethasone, VEGF and TSP-1 were also tested [137]. Intriguingly, a significant correlation between high plasma levels of VEGF and high PSA values (r = 0.4223; P < 0.001) was observed. VEGF levels were moreover significantly increased in non-responders patients.
Table 5

Clinical studies concerning the use of chemotherapy with a metronomic schedule in prostate cancer

Metronomic protocol

Study design

Histology

Patients (N)

PSA reduction ≥50 % (%)

Symptoms relief (%)

Median PFS (months)

Median OS (months)

Reference

mCTX

Phase II

HRPC

58

17

 

7^

Not reached

Lord [133]

mCTX-Dex

Retrospective

HRPC

32

69

 

9*

 

Glode [134]

mCTX-Dex

Pilot

HRPC

17

23

30

3*

24

Nelius [135]

mCTX-Prednisolone

Phase II

HRPC

23

26

43

6

11

Ladoire [136]

mCTX-Dex-CXB

Phase II

HRPC

28

32

80

3

21

Fontana [137]

mCTX-UFT-Extramustine

Phase II

HRPC

21

57

10

7*

 

Nishimura [138]

mCTX-mMTX

Phase II

HRPC

58

25

30

5

11

Gebbia [139]

mCTX-UFT-Dex

Retrospective

HRPC

57

63

69

7*

 

Hatano [140]

PSA prostate-specific antigen, PFS progression-free survival, OS overall survival, HRPC hormone-refractory prostate cancer, mCTX metronomic Ciclophosphamide, DEX dexametasone, CXB Celecoxib, mMTX metronomic Metotrexate

^ Median duration of response; * median time to progression

In another study, patients with HRPC received oral CTX, UFT and estramustine phosphate continuously until disease progression or unacceptable toxicity [138]. The authors reported moderate efficacy and good tolerability.

Metronomic CTX was also associated with low-dose MTX in patients with metastatic HRPC, who progressed after docetaxel therapy [139]. Three patients (18 %) obtained PR and four SD (23 %).

Moreover, in a recent study, the effects of a combination of dexamethasone, UFT and CTX orally delivered, were retrospectively regarded in 57 patients with HRPC [140]. Two out 7 patients (28 %) with a measurable disease achieved PR and 5 SD (72 %).

The bulk of data supporting the efficacy of metronomic CTX allows to consider this approach a suitable option for patients with HRPC even if refractory to other cytotoxic drugs [141].

Renal cell carcinoma

Systemic treatment for renal cell carcinoma has lately had a huge impulse by the development of targeted therapies. Cytokines-based treatments, considered the only active therapies in the past, have been exceeded by newer molecules. Sunitinib, sorafenib, pazopanib, temsirolimus, everolimus, along with bevacizumab, all concur to control disease progression both in naive and pre-treated patients diagnosed with metastatic renal cell carcinoma (MRCC). The multikinase inhibitor sunitinib, as a first-line treatment for predominantly clear-cell renal carcinoma, prompted a gain of about 6 months in PFS compared to interferon-α (11 vs. 5 months) and an ORR of 31 % [142]. Similar results along with a good tolerability profile have been achieved with the use of pazopanib, an oral angiogenesis inhibitor targeting VEGFR1-3, PDGFR α and β and cKIT [143]. The comparison of sorafenib (another multikinase inhibitor) [144] or everolimus (an inhibitor of mTOR) [145] with placebo, as second-line therapy, has shown an improved PFS in the treatment arms (5.5 vs. 2.8 months in the first study and 4.0 vs. 1.9 months in the second one). Moreover, axitinib, a second generation of VEGFR 1–3 inhibitor, in renal clear-cell carcinoma patients progressed after sunitinib, determined even better results than sorafenib [146]. In such expanding and targeted therapy-oriented scenario, only limited data about MCT in MRCC are available (Table 6). On the one hand, a pioneering study investigated the pleiotropic agent, thalidomide, administered in a low-dose schedule in 66 patients with different advanced malignancies, including ovarian cancer (OC) and MRCC [147]. All patients experienced CB, while only a minority with MRCC obtained PR. On the other hand, a combination of metronomic CTX and CXB induced really disappointing results in a successive trial with a 12 % of CB [148].
Table 6

Clinical studies concerning the use of chemotherapy with a metronomic schedule in renal cell carcinoma and ovarian cancer

Metronomic protocol

Study design

Histology

Patients (N)

ORR (%)

Clinical benefit (%)

Median PFS (months)

Median OS (months)

Reference

mThalidomide

Phase II

MRCC

18

17

34

  

Eisen [147]

OC

19

0

5

  

mCTX-CXB

Phase II

RCC

32

3

12

3

14

Krzyzanowska [148]

Pioglitazone-Etoricoxib-IFN-α-mCAP

Phase II

MRCC

45

35

 

7

27

Walter [149]

mCAP-GEM-Sorafenib

Phase II

MRCC

40

50

92

11

26

Bellmunt [150]

mCTX-Bev

Retrospective

OC

15

53

73

6

 

Chura [153]

mCTX-Bev

Retrospective

OC

38

40

49

4

11

Sanchez-Munoz [154]

mCTX_Bev

Phase II

OC

70

24

87

7*

17

Garcia [155]

ORR overall response rate (complete response + partial response); clinical benefit (complete response + partial response + stable disease), PFS progression-free survival, OS overall survival, m metronomic, mCTX metronomic Ciclophosphamide, CXB celecoxib, mCAP metronomic Capecitabine, IFN-α: interferon-α, GEM gemcitabine, Bev bevacizumab

* Median time to progression. MRCC metastatic renal cell carcinoma, OC ovarian cancer

A phase II study evaluated the impact of an anti-inflammatory treatment associated with reduced dose of capecitabine in MRCC [149]. To this purpose, a combination of etoricoxib, pioglitazone, interferon-α and capecitabine (dose slightly lower than the standard one) was used. The ORR was far from negligible (35 %); moreover, a strong CRP decline was observed in all patients with elevated basal-CRP levels.

More recently, in MRCC, the activity of a multitarget chemo-switch regimen has been tested as first-line therapy [150]. Forty patients received metronomic capecitabine associated with standard doses of both gemcitabine and sorafenib, for six cycles, followed by sorafenib monotherapy. Response rate and PFS were higher than those previously observed with gemcitabine and capecitabine or sorafenib monotherapy. However, toxicity was relevant and a death due to pulmonary embolism was reported.

The few existing data support a synergistic effect of MCT and antiangiogenic agents in MRCC.

Ovarian cancer

Frontline chemotherapy for ovarian cancer usually includes taxanes and platinum compounds. The adjunct of bevacizumab to conventional chemotherapy improved PFS relative to chemotherapy alone (14.1 vs. 10.3 months, HR: 0.71, P < 0.001) but not OS [151]. Platinum-sensitive recurrence (i.e. beyond 6 months from the treatment’s end) can be re-treated with a platinum-based combination. In platinum-resistant disease (i.e. recurrence within 6 months), different agents like docetaxel, oral etoposide, gemcitabine, liposomal doxorubicin, weekly paclitaxel and topotecan may be employed. However, the ORR and OS achievable with these treatments appear absolutely low, not exceeding 20 % of the cases and 12 months, respectively [152].

Various reports have documented clinical activity of combining bevacizumab plus metronomic CTX. This pair is credited with a synergic antiangiogenic effect (Table 6). Metronomic CTX in association with bevacizumab was firstly investigated in 15 patients with recurrent OC [153]. Despite patients were heavily pre-treated (≥5 regimens), the RR was 53 % without significant toxicity. Moreover, another retrospective study, adopting the same chemotherapy schedule, reported similar effects [154].

Differently, a larger phase II trial, including 70 patients with heavily pre-treated OC treated with the same drugs combination, did not show satisfactory results [155]. CB was good (87 %), but the ORR was lower while bevacizumab-related toxicities were considerable. Only one schedule of metronomic CTX has been explored in OC so far, with conflicting results, so no definitive conclusions can be drawn on the activity of metronomic approach in OC.

Discussion and conclusions

MCT is currently prescribed by Oncologists in different types of malignancies, mostly outside a clinical trial and after failure of at least two previous lines of treatment [156]. Unlike MTD therapies, doses of which were established in well-defined phase I studies, both doses and frequency of metronomic protocols have been generally empirically devised in an attempt to determine the optimal biological dose (OBD): that is the dose causing maximum reduction of tumour volume with none or minimal toxicity [157]. Tumour angiogenesis is recognised as the basic target of MCT; nevertheless, accumulated evidence showed that more than an antiangiogenic therapy, MCT is a form of multitarget cancer therapy. Indeed, recent pre-clinical and clinical studies have shown that MCT induced significant immunomodulatory effects, such as depletion of Tregs within the tumour microenvironment [158160]. A number of additional mechanisms have also emerged, including reduction in cancer cell stemness [28] and selective inhibition of HIF-1α by metronomic doxorubicin [20, 21].

In a pioneering study, Browder and co-workers demonstrated metronomic CTX was more effective compared with the conventional schedule and capable of preventing drug resistance in a drug-sensitive tumour [4]. MCT schedules opened the fascinating prospective of overcoming drug resistance allowing re-use of traditional cytotoxic drugs.

A genetic stability of vascular endothelial cells has been retained to prevent chemo resistance during metronomic protocols. Nevertheless, tumour endothelial cells have also shown the disposition towards tumour-specific genetic abnormalities [161, 162] and the acquisition of drug resistance [163]. However, it was recently shown that during repeated exposure to MCT chemotherapy, there is an increased chemosensitivity of the endothelial cells [164]. Moreover, during MTD chemotherapy-free intervals, as part of an adaptive response to the chemotherapy-induced myelosuppression, there is a conspicuous release of hematopoietic progenitors, including CEPs, from the marrow into the peripheral blood circulation. CEPs release is considered a pivotal step in the process of accelerating tumour re-growth and drug resistance. It was shown that in immune-deficient mice-bearing human lymphoma cells, the delivery of metronomic CTX did not mobilize CEPs differently from MTD CTX [165]. Finally, MCT could also act directly and continuously on tumour cells [166], thus avoiding the rebound in tumour cell division known as “repopulation” [5]. It has been reported that this phenomenon, occurring during the chemotherapy-free intervals, worsens with continued treatment causing the tumour to become very aggressive [167].

Since MCT is a slow acting cancer therapy and might be devised as a personalised treatment, it would require predictive biomarkers of activity and response. It was demonstrated in pre-clinical models that OBD of different metronomic regimens cause a decrement of CEPs. These experiments were considered proof of concept that MCT acts on the vasculature. Consequently, the measurement of CEPs in peripheral blood, as a valuable surrogate biomarker, has been proposed to assess the OBD of different metronomic schedules in clinical studies. Low baseline CEPs were shown to be predictive of both better PFS [71] and clinical benefit in patients with solid tumours [168]. Accordingly, high basal level of CEPs was a significant predictor of poor PFS and OS in HCC patients treated with MCT plus Sorafenib [169]. Moreover, CEPs increased significantly over 2 months in patients with no clinical benefit [168].

CD133, which is expressed on primitive cells including CEPs, has been also hypothesised as a possible predictive biomarker of poor response to MCT [120]. In fact, CD133/prominin-1 ratio was substantially raised in patients with progressive disease after treatment with a metronomic CTX, UFT and CXB, thus implying a possible increase in CEPs. In spite of interesting data emerging from clinical studies, the use of CEPs as biomarker needs to be further elucidated; in fact, many controversies still regard their own identification [170].

Circulating endothelial cells (CECs), reflecting an active vascular remodelling in cancer vessels, were also used as predictive biomarker in breast cancer patients receiving MCT plus Bevacizumab [70, 171]. High baseline CECs levels were significantly associated with both the achievement of CB and increased PFS [70, 171]. A pattern of decreased CECs along with increased angiogenic growth factors at the time of disease relapse was also described, probably reflecting a switch towards a novel type of cancer vascularization [171]. Moreover, CECs were also shown to be reduced compared with pre-treatment values in patients with advanced slow-growing solid tumours, who achieved CB after metronomic treatment [168]. Therefore, it was shown in different reports that a high baseline CECs count and changes in CECs levels might be, respectively, useful to select MBC patients who would benefit from MCT plus bevacizumab [70, 171] and to predict response to different metronomic treatments [70, 168]. However, there were also contradictory data: CECs and CEPs level in fact did not change during treatment for advanced malignancies with a diverse metronomic schedule [172].

VEGF-A, used as a surrogate biomarker of efficacy also, generated discrepant results. It was reported that VEGF-A concentrations rose immediately after chemotherapy administration [120], possibly reflecting hypoxia of tumour tissue, dropping to lower levels during subsequent chemotherapy cycles. The drop of VEGF levels during MCT administration predicted tumour response in several studies [120, 171]. Conversely, other authors did not report significant differences in VEGF-A concentrations between responders and non-responders to a metronomic approach in diverse tumours [53, 54, 57, 87, 114].

A pre-clinical study, investigating several metronomic regimens, demonstrated TSP-1 acts as a mediator of the antiangiogenic effects [173]. Moreover, a phase I dose-finding trial showed TSP-1 plasma level and its gene expression in peripheral blood mononuclear cells markedly increased during administration of metronomic irinotecan. Worth of note, the highest increase was at the lowest drug dose level [117]. Following data also supported TSP-1 level as a possible surrogate biomarker for MCT activity [120, 174]. TSP-1 levels were significantly higher in SD than in PD patients with refractory gastrointestinal cancer treated with a metronomic schedule of CTX, UFT and CBX [120]. Conflicting results were on the other hand reported by another group in heavily pre-treated patients receiving CTX-based MCT [175]. Given the small sample size of most studies and the dispersion of the results in both different types of cancer and treatments, we still need to standardize their evaluation methods and confirm the usefulness of proposed biomarkers in larger randomized trials. The ideal biomarker of MCT should allow to candidate patients for MCT, to select the most effective schedule and control treatment effectiveness over time.

Almost all cytotoxic drugs have been experimented in low continuous doses but CTX is by far the most widely used. Its effectiveness was demonstrated in a randomized trial in progressive and advanced cancer patients having exhausted all effective standard therapies [176]. Combinations of metronomic CTX with either capecitabine [62, 63] or MTX [54, 55] were shown to confer CB in about half of pre-treated MBC patients. The addition of an antiangiogenic molecule such as bevacizumab further ameliorated the treatment outcome [7072]. Furthermore, in HRPC patients, alone or combined with other drugs, CTX allowed both a marked decrease in PSA levels as relevant relief of symptoms [136140]. Positive results were achieved even in heavily pre-treated [55, 62, 64, 72, 134, 176] and in elderly patients [67, 99, 105, 127]. While in paediatric patients with refractory/relapsed rabdomyosarcoma, strikingly good results were reported in a large phase 2 study [101]. Metronomic schedules of fluoropyrimidines also provided interesting results both in breast [64, 65] and in gastrointestinal cancer [118, 120]. On the other hand, several studies reporting on metronomic schedules of alkylant agents, topoisomerase inhibitors and antitubulin compounds yielded disappointing results in tumours considered poorly responsive to traditional chemotherapy and with a dismal prognosis such as gliomas [34, 35], melanoma [58, 109] or MRCC [148]. This was despite they had been coupled with other antiangiogenic agents [35, 36, 41] and in spite of the high vascularization of such tumours which provides the best rationale relative to the use of an antiangiogenic treatment. Both primary and acquired resistance mechanisms to antiangiogenic therapy could underlie such dismal results [177]. Moreover, since the therapeutic window of MCT may be narrow [157], several metronomic schedules do not actually fulfil a truly MCT approach but consist in just a little reduction of the standard MTD doses.

Pre-clinical data demonstrated metronomic schedules are particularly suited for the association with new targeted therapies [6, 17]. Encouraging clinical results have already been achieved by combining different metronomic protocols with bevacizumab or gefitinib in NSCLC [85, 90], MBC [7072] and recurrent OC [154]. Other valuable studies on MCT have been aimed at experimenting a chemo-switch strategy: cytotoxic drugs delivered at metronomic dosage after or within cycles of MTD chemotherapy. In this approach, MTD chemotherapy is delivered for tumour debulking, while multitargeted, antiangiogenic maintenance regimen follows to avoid the inevitable rebound and regrowth (resistance), of tumour cells [17, 24, 74, 120, 122, 150]. These schedules yielded good results in paediatric brain tumours [24, 45], in GI cancer [120, 122] and in MRCC [150].

It has been shown that in vitro, intermittent drug deprivation affects cancer cells which have got used to survive on continuous drug administration [178180]. From these experimental data, it has been theorized the 4D effect [18]. Proof-of-concept clinical trials showed high activity of schedules which exploit this mechanism [25, 45, 181].

Good tolerability, contributing to improve patients quality of life, is a pivotal feature of metronomic regimens. It has been suggested growth arrest and/or apoptosis of endothelial cells, during MCT, is mediated by endogenous inhibitors of angiogenesis, which do not affect hematopoietic progenitors [182].

The majority of clinical studies shows these schedules are generally well tolerated causing only mild toxicity [53, 120]. Grade 3 toxicities in fact were rarely reported and included nausea and/or vomiting, hypertransaminasemia, fatigue and neutropenia [55, 57, 58, 70]. However, when MCT was combined with targeted agents such as bevacizumab or sorafenib, more severe toxicities were observed [150]. Another concern is with regard to long-term toxic effects especially in children. Most MCT schedules have been delivered to adults with a brief life expectancy, for this reason potential long-term toxic effects are mostly unknown. MCT generally consists in prolonged administration leading to high total cumulated doses of anticancer agents. High cumulated doses of temozolomide and etoposide have been shown to cause myelodisplasia and secondary leukaemias [183, 184]. Physiologic angiogenesis mechanism may differ from tumour neo-angiogenesis [18]. Nonetheless, special attention must be given to the development of young children receiving metronomic chemotherapy since angiogenesis plays an important role in growth.

Overall, the results of MCT schedules are highly variable ranging from satisfactory to absolutely disappointing. It should be emphasized many studies were retrospectively conducted and reported on a wide variety of schedules in heterogeneous tumour types and patients setting. Few clinical MCT studies were instead prospectively conducted and included a relevant number of homogeneous patients and only a minority were randomized [56, 69, 76, 90, 110]. This lack of homogeneity often hampers the identification of most effective MCT schedules in diverse tumour settings.

We believe MCT schedules are active and very promising in diverse tumour types. The combination of MCT protocols with newer targeted molecules opens unexplored and exciting scenarios nonetheless, in order to move firmly forward, much work still awaits to be done in clinical trials outline. Studies comparing outcomes of the best metronomic schedules versus standard chemotherapy in fit patients or with the best supportive care in heavily pre-treated or frail patients are in fact awaited. Furthermore, well-designed clinical trials would also be the ideal standing for exploring and selecting truly reproducible biomarkers.

Copyright information

© Springer-Verlag Berlin Heidelberg 2013