Main

Metronomic chemotherapy refers to the frequent, usually daily, administration of cytotoxic drugs at doses that are significantly less than the maximum-tolerated dose, with no prolonged drug-free breaks. Oral cyclophosphamide-based metronomic chemotherapy is the most largely studied metronomic regimen with more than 30 retrospective studies and phase II trials reporting in vivo antiangiogenic and immunomodulatory properties and significant clinical antitumour activity, which has been confirmed in heavily treated patients that have exhausted all effective treatments (Penel et al, 2010,2012).

Imatinib mesylate (IM) is a potent inhibitor of KIT, ‘break-point cluster region’-‘Abelson’ (BCR-ABL), platelet-derived growth factor (PDGFR) and colony-stimulating factor 1 receptor (CSF1R) activities that has marked antitumour activity on chronic myelogenous leukaemia and several other haematological malignancies and solid tumours including gastrointestinal stromal tumours (GISTs) (Blay et al, 2007), chordoma (Stacchiotti et al, 2012), aggressive fibromatosis (Penel et al, 2011a) and melanoma harbouring KIT mutation (Guo et al, 2011), as well as thymic carcinoma with KIT mutations (Schirovi et al, 2012).

Pietras and Hanahan (2005) described the synergistic activity of the combination of metronomic cyclophosphamide (MC) and IM in patient-derived tumour xenograft mouse models. Their study showed that this combination acts as an antiangiogentic treatment, as imatinib interferes with pericyte activity through the inhibition of the PDGFR and oral MC induces the endothelial cell apoptosis. This combination provides an objective response and survival advantage in this mouse model. Until now, the precise antitumour activity of this combination had not precisely determined in humans despite these promising preclinical findings.

In this context, a multi\e dose-escalating phase I trial to determine the phase II recommended dose of this combination, to determine the impact of oral MC on the IM pharmacokinetic (PK) and to seek some signs of activity in patients with IM and sunitinib-refractory GIST, with chordoma, life-threatening aggressive fibromatosis or melanoma harbouring KIT mutation was conducted.

Materials and methods

Study design

This is a 3+3 dose-escalation study, the primary objective of which is to determine the maximum-tolerated dose of a combination of IM associated with fixed dose of oral MC. The eligibility criteria are listed in Table 1. Tumour response was assessed by RECIST 1.0 every 8 weeks (Therasse et al, 2000).

Table 1 Eligibility criteria
Full size table

Dose-escalating process and definition of the maximum-tolerated dose

Eligible patients received dose-escalated IM (three dose levels: 400 mg daily, 300 mg two times daily, 400 mg two times daily) for 15 days (until steady state) followed by continuous 50 mg day of MC associated with IM. The choice of MC dose was based on widespread evidence of the long-term tolerability of this schedule (Blay et al, 2007; Penel et al, 2011a). The dose levels of IM have been chosen because of the well-established long-term tolerability of this drug.

Dose-limiting toxicities (DLTs) were assessed during the first 6 weeks of treatment and included an interruption of treatment for more than 4 consecutive days during the first 6 weeks, and the following toxic events (NCI-CTCAE version 3): prolonged (>7 days) grade 4 neutropenia, febrile neutropenia, grade 4 thrombopenia, haemorragia related to thrombopenia, grade 3–4 diarrhoea, grade 3–4 rash as well as any kind of unexpected life-threatening toxicities.

We planned an expanded cohort of 14 additional patients at the phase II recommended dose to better explore the tolerability and the activity of this combination (Penel et al, 2011b).

Pharmacokinetic analyses

Pharmacokinetic analyses were carried out by EC and MM at the Pharmaco-Toxicology Laboratory (University Hospital, Bordeaux, France). Samples had been collected at day 15 (IM alone) and at day 28 (IM+MC). Pharmacokinetic analyses were scheduled before (until steady state) and during the combination at T0 (predose), 30 min, 1 h, 2 h, 3 h, 4 h, 6 h, 10 h, 11 h, 12 h (day 1) and 24 h. Metronomic cyclophosphamide and IM were quantified in the serum by liquid chromatography-tandem mass spectrometry methods validated according to Federal Drug Agency guidelines. The mass spectrometer was operated in electrospray ionisation mode with multiple reaction monitoring. Samples (300 μl) were prepared by SPE using an MCX Oasis microelution plate (Waters Corporation, Milford, MA, USA), after the addition of an acidic solution (pH 3) containing Internal Standard. The same solid-phase extraction procedure was applied for both methods, but samples were then analysed on different systems: IM was quantified using ultra-performance liquid chromatography coupled with a tandem quadrupole detector (Waters), whereas MC samples were analysed on an Alliance 2695 coupled with Quattro Micro (Waters) (Bouchet et al, 2011).

Ethical considerations

This study was approved by the regional Ethics Committee (‘Comité de Protection des Patients Nord-Ouest III’, date of approval: 29 May 2009) and by the French Health Products Safety Agency (‘Agence Française de Sécurité Sanitaire et des Produits de Santé’, 1 June 2009). This study was registered in the European Clinical Trials Register (EudraCT No. 2008-004212-12). Informed consent was obtained from each patient.

Results

Description of the population

Twenty-six patients were included between February 2009 and January 2011. During the dose-escalation phase, five patients (the first two patients had been replaced because they were not assessable for the PK analysis) have been enrolled at first dose level (IM 400 mg per day), three patients at second dose level (IM 300 mg two times daily) and four patients at the third dose level (IM 800 mg per day). One of the patients enrolled in the third dose level was never treated (died at day 13 days of biguanide overdose). Fourteen additional patients were enrolled in the expanded cohort (IM 800 mg per day). The study population consisted of 20 men (77%) and six women (23%). The median age was 60.5 (range, 41–80) years. The primaries were GIST (17, 65%), sacral chordoma (7, 27%) and mucosal melanoma (2, 8%). The primary locations of GIST were as follows: small intestine (9), stomach (5), rectum (2) and mesenteric (1). Twenty patients (77%) had metastatic disease, mainly involving the liver (16 patients), lung (11 patients) or peritoneum (8 patients). At study entry, the performance status (PS) was as follows: PS=0 in 11 patients (42%), PS=1 in 14 patients (54%) and PS=2 in 1 patient (4%). The previous treatments were surgery in 17 cases (65%), radiotherapy in 5 cases (19%) and previous molecular targeted therapies in 19 cases (73%). One of the two patients with mucosal melanoma had received previous chemotherapy with dacarbazine. Two of the seven patients with chordoma had received previous treatment with IM alone. All patients with GIST have previously been treated with IM alone and suntinib and had progressed. The other treatments administered to patients with GIST were as follows: sorafenib (11 patients), nilotinib (6 patients), IM+nilotinib (2 patients), masatinib (2 patients) and motesanib (1 patient). All patients experienced progressive disease at study entry.

Dose escalation

No DLT was observed during the dose-escalation phase: the third dose level was retained as the recommended dose for further phase II trial: MC 50 mg two times daily plus IM 800 mg per day.

Tolerance

One patient enrolled in the expanded cohort was not assessable for safety (died at day 13 of biguanide overdose). The safety profile is detailed in Table 2. Treatment had been suspended for toxicity reason in seven cases (6 out of 14 patients treated at the phase II recommended dose). Doses were reduced for 4 out of 14 patients treated at the phase II recommended dose. One case of acute myeloid leukaemia in a chordoma patient treated with this association for 4 years, thus, a causal relationship is possible.

Table 2 Maximum treatment-related toxicity (n=25)
Full size table

Antitumour activity

Twenty-five patients were assessable for activity. No objective response was observed during the study. The median progression-free survival (PFS) was 3.4 months (95% CI: 2.5–9) and the median overall survival (OS) was 13.4 months (95% CI: 3.0–16.0+) in the entire study population. These figures varied widely according to the tumour types. Imatinib mesylate/suntinib GIST patients experienced a PFS of 2.9 months (95% CI: 2.3–7.6) and an OS of 11.0 months (95% CI: 3.0–16.0). Chordoma patients experienced long-lasting stable disease, whereas all had progressive disease at study entry. The median PFS and median OS were not reached for chordoma patients. The PFS and OS of mucosal melanoma patients were 2.9 and 5.2 months, respectively (see Figures 1 and 2).

Figure 1
figure 1

Progression-free survivals by tumour types.

Full size image
Figure 2
figure 2

Overall survivals by tumour types.

Full size image

Pharmacokinetic

The area under the curve for IM correlated to the administered dose (Figure 3). Metronomic cyclophosphamide does not modify the PK parameters of IM (Table 3).

Figure 3
figure 3

Area under curve (AUC) of imatinib according to the three dose levels: 400 mg daily, 300 mg two times daily, 800 mg daily).

Full size image
Table 3 PK parameters of imatinib
Full size table

Discussion

The key findings of this dose-escalating phase I trial were: (i) the phase II recommended dose is MC 50 mg two times daily plus IM 800 mg per day, (ii) this schedule is well tolerated, (iii) MC does not interfere with the IM PK and (iv) tumour growth arrest was observed in patients with imatinib/sunitib-refractory GIST and chordoma, requiring further clinical investigations.

The tolerance of the combination was mostly manageable without unexpected toxicity. Arterial hypertension and renal toxicity (proteinuria/microscopic haematuria) were only observed in one case. The case of acute leukaemia is possibly related to this combination; cyclophosphamide is a well-established carcinogen (Xu et al, 2013). This patient with a massive sacral chordoma was one of the first enrolled patients and received 400 mg per day of IM plus MC after failure of IM alone. This combination provided pain relief and alleviated nerve palsies. After 18 months of treatment, we proposed a drug holiday, but this drug holiday led to severe worsening of pain and nerve palsies. Therefore, this patient received this combination for 5 years (from February 2009 to February 2013) without significant toxicity until the diagnosis of acute leukaemia. This patient is currently under treatment for this leukaemia.

We have observed clinically meaningful activity in seven chordoma patients. The median follow-up of the chordoma patients was 23.7 months. The 12-month OS rate was 6/7 and the 12-month progression-free rate was 3/7. The median PFS was not reached. Stachiotti et al (2013) have recently reported on a single-arm phase II trial of IM (800 mg per day) in 56 chordoma patients. The objective response rate was 1 of 50 (2%). The median PFS was 9 months (Stachiotti et al, 2013). The same team had explored the activity of lapatinib (1500 mg per day) in 18 EGFR-positive chordomas. The median PFS was 8 months, with 6 confirmed partial responses (Stachiotti et al, 2013). The IM/MC combination warrants further investigation in chordoma patients.

Seventeen previously heavily treated GIST patients have been enrolled in this trial. In these cases, the PFS was 2.9 months. Growth modulation index (ratio of time to progression (TTP) with the investigational combination divided by the time to progression with the prior treatment) was available in 14 out of 17 GIST patients. The median TTP was 8.2 months (1.2–31.2) with the prior treatment and 2.7 months (0.5–10.5) with the IM/MC combination. Three out of the 14 available patients (21%) experienced a significant growth growth arrest with a growth modulation index superior to 1.33 (Von Hoff et al, 2010). These figures must be compared with the PFS observed in large retrospective study. The rechallenge of IM alone provides a median PFS of 2.9 months. The role of adding MC to rechallenge of IM in heavily pretreated patients have to be studied in an appropriate randomised trial enrolling selected cases (e.g., GIST with multiple mutations or GIST with D842V mutation).

This study demonstrates that adding MC does not interfere with the PK parameters of IM. Up until now, most trials did not document PK interaction in cases of molecularly targeted agent/cytotoxic agent combinations (Giaccone et al, 2004; Messersmith et al, 2004; Thienelt et al, 2005; Sweeney et al, 2010; Bousquet et al, 2011; De Jonge et al, 2011; Awada et al, 2013). In a phase I trial, Hamberg et al (2010) have found that exposure to ifosfamide (another oxazaphosphorine, such as cyclophosphamide) decreased the exposure to sunitinib. In this trial, the schedule of ifosfamide was conventional (6–9 g m−2 per 3–4 weeks). In the present trial, the administered dose of cyclophosphamide is very low (50 mg two times daily) in comparison to the doses of cyclophosphamide that have been administered in a metronomic manner.

We have explored the safety of this association, observed some signs of activity and measured the impact of adding MC on IM PKs parameters. This combination warrants further clinical exploration in chordoma patients and, to a lesser extent, in heavily pretreated GIST patients.