¹Clinical Hematology Unit, CHU Le Bocage, Dijon, France

²INSERM U517, Faculty of Medicine and Pharmacy, Dijon, France

³Ambulatory Blood Pressure Monitoring Unit, CHU Le Bocage, Dijon, France

⁴Cardiology Unit, CHU Le Bocage, Dijon, France

⁵Clinical Hematology Unit, CHU Claude Huriez, Lille, France

⁶Pharmacology Unit, Institut Claudius Regaud, Toulouse, France

⁷Debiopharm SA, Lausanne, Switzerland

⁸Oncodesign SA, Dijon, France

Correspondence to: E Solary, Clinical Hematology Unit, CHU Le Bocage, BP1542, 21034 Dijon Cedex, France; Fax: 33 3 80 29 50 42

Overexpression of P-glycoprotein (P-gp) in cancer cells reduces intracellular accumulation of various anticancer drugs including anthracyclines and vinca alkaloids. This multidrug resistance (MDR) phenotype can be reversed in vitro by a number of non-cytotoxic drugs. We have identified the quinine's isomer cinchonine as a potent MDR reversing agent, both in vitroand in animal models. Here, we report an open phase I dose escalation trial in patients with refractory or relapsed malignant lymphoid diseases. Cinchonine dihydrochloride was administered by continuous i.v. infusion for 48 h and escalated over five dose levels ranging from 15 to 35 mg/kg/d. Cinchonine infusion started 24 h before i.v. doxorubicin (25 mg/m²), vinblastine (6 mg/m²), cyclophosphamide (600 mg/m²) and methylprednisolone (1 mg/kg/d) (CHVP regimen) and lasted for 24 h after chemotherapy infusion. Thirty-four patients received 87 cycles of CHVP/cinchonine. The MTD of cinchonine administered by continuous i.v. infusion was 30 mg/kg/d. Prolonged cardiac repolarization was the main dose-limiting toxicity. No ventricular arrhythmia including 'torsade de pointes' was observed. An MDR reversing activity was identified in the serum from every patient and correlated with cinchonine serum level. When infused at 30 mg/kg/d, cinchonine demonstrated a limited influence on doxorubicin pharmacokinetic. We conclude that i.v. infusion of cinchonine might be started 12 h before MDR-related chemotherapy infusion and requires continuous cardiac monitoring but no reduction of cytotoxic drug doses. Leukemia (2000) 14, 2085-2094.

cinchonine; drug resistance; P-glycoprotein; phase I clinical trial; lymphomas; myelomas

Introduction

Resistance of malignant cells to cytotoxic drugs is one of the main reasons for the failure of chemotherapy. Experimental systems including in vitro cultured cell lines have identified several mechanisms of cellular resistance to chemotherapeutic agents.¹ One of these mechanisms is the so-called 'multidrug resistance' (MDR) that involves an increased expression of the mdr1 gene product, a 170 kDa glycoprotein named P-glycoprotein, in tumor cell plasma membrane. This energy-dependent transmembrane pump, that belongs to a large superfamily of highly conserved ATP-binding cassette proteins, facilitates the cellular efflux of various substances, thereby reducing their intracellular concentration.² In normal cells from various tissues, eg biliary canaliculi,³ endothelium of the blood-brain barrier⁴ and bone marrow stromal cells,⁵ P-glycoprotein is thought to act as a detoxifying agent by pumping toxins or xenobiotics out of these cells. In organs such as adrenal glands, P-glycoprotein might be involved in the transport of steroid hormones.⁶ Overexpression of P-glycoprotein in cancer cells reduces intracellular accumulation of a broad array of anticancer drugs including anthracyclines, vinca alkaloids, and epipodophyllotoxins.² Expression of mdr1 mRNA and the presence of P-glycoprotein have been detected in a large number of human tumors, including malignant lymphomas,^7,8,9,10 multiple myelomas^8,11 and chronic lymphocytic leukemias.¹² Moreover, MDR phenotype of tumor cells has been correlated with poor response rates to chemotherapy as well as shortened event-free survival and overall survival in a variety of malignancies.^{1,13,14,15,16}

The P-glycoprotein-mediated decrease of intracellular accumulation of cytotoxic agents can be reversed efficiently in vitro by a number of non-cytotoxic drugs.¹⁷ The clinical use of most of these drugs is precluded by serum protein binding or clinical toxicity.¹⁸ We observed that serum from patients receiving conventional doses of quinine by intravenous infusion reversed the MDR phenotype of rat colon cancer¹⁹ and human leukemic cells.²⁰ Based on this observation, quinine appeared as a good candidate for evaluating the clinical interest of MDR reversing agents. Quinine was demonstrated to be safe when combined with mitoxantrone and cytarabine for treating acute leukemia patients²¹ as well as with doxorubicin and vincristine for treating patients with refractory multiple myeloma or malignant lymphomas (unpublished data). Although quinine did not significantly improve the response rate of refractory and relapsed acute myelogenous leukemias,²² we demonstrated its ability to improve the response of P-glycoprotein-expressing myelodysplastic syndromes to a mitoxantrone-cytarabine combination.²³

In a search for a more effective MDR inhibitor, we have identified cinchonine as the most potent cinchona bark alkaloid, both in vitro and in animal models.²⁴ Cinchonine is more efficient than quinine in increasing the intracellular accumulation and restoring the cytotoxicity of doxorubicin, mitoxantrone and vincristine on well-characterized MDR cell lines.²⁵ The plasma proteins have limited influence on the MDR reversing activity (MRA) of cinchonine and the fraction of the compound that is trapped in red blood cells is rapidly and completely exchangeable with plasma.²⁶ In rats bearing resistant tumors, cinchonine is more efficient than quinine in increasing the efficacy of anticancer drugs.²⁵ Although cinchonine has been used previously in the clinic, either alone or in combination with quinine and/or quinidine for the treatment of malaria, its pharmacokinetics remained incompletely studied,^27,28 its tolerance in humans had not been widely investigated and its combination with cytotoxic drugs remained to be tested.

The present phase I study was undertaken to determine the maximal tolerated dose (MTD) of cinchonine administered by continuous intravenous infusion for 48 h in combination with cyclophosphamide, doxorubicin, vinblastine and prednisone (CHVP regimen) in patients with a refractory or relapsed lymphoproliferative disease. The main adverse effect of cinchonine was observed to be a dose-dependent increase of cardiac repolarization, as demonstrated by QT interval prolongation.

Patients and methods

Patients

The initial design of this open multicenter phase I trial and the subsequent amendments were approved by the institutional review board of Dijon hospital. Before treatment, every patient gave informed consent after having been advised about the purpose and investigational nature of the study as well as potential risks. Patients eligible for the study were those older than 17 and younger than 76 years of age with a histologically established diagnosis of malignant lymphoid disease, including non-Hodgkin and Hodgkin lymphoma, chronic lymphocytic leukemia and multiple myeloma. These patients were refractory to conventional chemotherapy including anthracyclines, or in relapse after having received at least one chemotherapeutic regimen including anthracyclines. Patients likely to be cured or to have prolonged response to conventional therapy were excluded. Those with limited life expectancy (<1 month), abnormal kidney function (creatinine >150 mol/l), abnormal hepatic functions (serum bilirubine >35 mol/l, transaminases or alkaline phosphatases >4 ´ N), bone marrow dysfunction unrelated to the disease or electrolyte disorder were ineligible. Other exclusion criteria included known hypersensitivity to quinine and related molecules, previous history of hemorrhagic cystitis after cyclophosphamide administration and abnormal cardiac function, including decreased isotopic left ventricular ejection fraction (<50%), arrhythmia requiring treatment and untreated atrio-ventricular heart block.

Regimen

Cinchonine dihydrochloride was supplied by Debiopharm SA (Lausanne, Switzerland) as a solution in water (50 mg/ml) and diluted in 5% glucose solution. As shown in Figure 1, cinchonine was given as a continuous i.v. infusion from day 1 to day 3, over a 48-h period. The first dose level was 15 mg/kg/d. The following steps were defined by a 5 mg/kg/d increase of cinchonine dose over the previous step. The chemotherapy regimen, that consisted of cyclophosphamide 600 mg/m² administered by 30-min i.v. infusion, doxorubicin 25 mg/m² as a 15-min slow i.v. bolus infusion and vinblastine 6 mg/m² as a 5-min i.v. infusion, was infused on day 2. Methylprednisolone (1 mg/kg/d) was administered as a 20-min i.v. infusion on days 2 and 3 and further given per os from days 4 to 6. Each patient received at least one complete cycle of treatment. This cycle was repeated every 28 days, depending on tolerance, response and patient's agreement, up to a maximum of six cycles. At least six patients were included in each step of the study. Patients required a central venous catheter for all infusions and were hospitalized during the first 4 days of the initial cycle of CHVP/cinchonine.

Monitoring of toxicity

Clinical symptoms, pulse rate and blood pressure were monitored twice a day by the investigator during cinchonine infusion. Laboratory tests (blood cell count, electrolytes, hepatic enzymes, urea and creatinine serum level) were performed before, during and after cinchonine infusion. Toxicity was assessed according to the NCI common toxicity criteria scale. At the beginning of the study, cardiac toxicity was assessed by one daily electrocardiogram (ECG) and continuous scope monitoring with pulse rate alarm. After having identified QTc prolongation and according to the recommendations of the Ethics board and the Agence Française du Médicament, patients were hospitalized in an intensive care unit. A Holter recording (see below) starting 4 h before cinchonine infusion and stopped 24 h after the end of this infusion was included in the monitoring procedure. In addition, five ECGs were recorded during every first cycle, with a double amplitude (2 cm/mV) and a paper speed of 50 mm/s on 12 standard leads with at least 10 complexes per lead. The five ECGs included two baseline registrations before treatment (period A), one during cinchonine infusion before chemotherapy (period B), one during cinchonine infusion, after the end of chemotherapy (period C) and one after the end of cinchonine infusion (period D) (Figure 2). All these ECGs were coded blindly and analyzed by an independent expert. Serious adverse events were recorded by the investigator who informed the sponsor within 24 h, according to the French legislation.

Holter analysis

ECG was recorded continuously on a Synesis digital recorder (ELA-Medical, Le Plessis-Robinson, France) equipped with a Flash card with the PCMCIA standard renewed every 24 h. Recordings were initiated at least 4 h before the start and stopped 24 h after the end of cinchonine infusion and stored on a 2.3 gigaoctet laser disk using a power MO2600 optical reading-recording device (Olympus Optical Co Ltd, Le Plessis Robinson, France). The standard Holter ELA-medical reading software (version 3.1) was used to recognize QRS complexes, to analyze heart rate variability in time and frequency domain, to identify arrhythmia and to measure the duration of QT intervals.

Response criteria

Clinical and biological parameters that characterized disease extension were registered at inclusion and checked after every chemotherapy cycle, together with the performance status. The main parameters used to evaluate patients with lymphoma were lymph node, spleen and liver enlargement, any other tumor localization and LDH serum level. In patients with CLL, we monitored the number of peripheral blood lymphocytes, hemoglobin serum level, platelet count and lymph node, spleen and liver enlargement. In patients with multiple myeloma, we checked immunoglobulin levels in serum and urine, bone marrow plasmocyte infiltration, creatininemia, calcemia, hemoglobin serum level, platelet count and bone X-rays.

Pharmacokinetics of cinchonine

Blood samples were collected at various times before, during and after cinchonine infusion. A 10 ml urine sample was collected from the 24-h urines, before treatment, after 24 and 48 h of cinchonine infusion and 24 h after the end of cinchonine infusion (Figure 1). Total cinchonine concentration in serum and urine samples was determined by using a previously described high performance liquid chromatography (HPLC) method.¹⁸ Briefly, 100 l of the tested sample was mixed with 100 l hydroquinidine (1 g/l; Sigma-Aldrich, L'Isle d'Abeau, France) as an internal standard. Extraction was performed with a 5 ml mixture of dichloromethane (Merck, Darmstadt, Germany) and isoamylic alcohol (98:2, v/v; Sigma-Aldrich). After centrifugation, the organic phase was evaporated to dryness, dissolved in the mobile phase (potassium dihydrogen phosphate 0.045 M, acetonitrile, 4:1, v/v, pH 3.8) and injected into the HPLC apparatus equipped with a fluorescent detector (Waters, Saint-Quentin, France). The stationary phase was a Lichrospher 60 RP-select B column (125 ´ 4 mm, 5 ) coupled with a pre-column Lichrospher 60 RP-select B 5 m column (4 ´ 4 mm). The detection was made by fluorometry at excitation and emission wavelengths of 315 and 418 nm, respectively. A calibration curve was used to determine the cinchonine concentration in each tested sample. The lower limit of detection was 0.01 g/ml and the lower limit of quantification was 0.2 g/ml.

Pharmacokinetics of doxorubicin

Plasma doxorubicin concentrations were also determined by the use of a HPLC method with fluorometric detection. Briefly, daunorubicin (10 g/ml) was added to the tested sample with 0.05 M borate buffer (pH 9.8) before extraction for 15 min with a 5 ml mixture of dichloromethane (Merck) and ethanol (4:1, v/v; Sigma-Aldrich). After centrifugation at 1000 g, the organic phase was evaporated to dryness, dissolved in the mobile phase (0.05 N H₃PO₄, acetonitrile, tetrahydrofurane, triethylamine, pH 2.5) and injected into the HPLC apparatus (Waters) equipped with a fluorescent detector (Shimadzu, Saint-Quentin, France). The stationary phase was a microBondapack Waters C18 column. The detection was made by fluorometry at excitation and emission wavelengths of 478 and 550 nm, respectively. A calibration curve was used to determine the doxorubicin concentration in each tested sample. The lower limit of quantification was 1 ng/ml and the last standard concentration was 5 ng/ml. The precision and accuracy for all tested concentrations, including the LOQ value (at 1 ng/ml) were within 15%.

Ex vivo bioassay measuring serum MDR reversing activity

To determine the ability of serum from cinchonine-treated patients to increase doxorubicin intracellular concentration by reversing the MDR phenotype, we used a previously described bioassay.²⁴ DHD/K12-TRb rat colon cancer cells, in which the MDR phenotype has been clearly identified, were seeded in microtiter plates (48 wells/plate) and cultured for 24 h, then incubated for 4 h at 37°C with 20 M doxorubicin (including 3% ¹⁴C-doxorubin and 97% non-radioactive drug) diluted in 0.35 ml serum-free HAM's F10 medium or serum samples obtained from patients. Serum-free medium containing 10 g/ml cinchonine hydrochloride was used as a positive control. After incubation, cells were rinsed three times with ice-cold phosphate-buffered saline (PBS), disrupted in 0.25 ml of 1 N NaOH and transferred into counting vials with 3 ml scintillant liquid (LKB, Stockholm, Sweden). The radioactivity, that represents intracellular doxorubicin accumulation (IDA) was measured on a scintillation counter (LKB 1241 Rackbeta, Stockholm, Sweden). Each determination was performed in duplicate with two wells per point. The serum reversing activity at time Tn was calculated according to the following formula: [(IDA at Tn - IDA at T0)/IDA at T0] ´ 100 where IDA is expressed as cpm.

Data analysis

The pharmacokinetic analyses were performed using the Siphar Software (SIMED, Chatenay-Malabry, France), following a non-compartmental model for cinchonine and a three-compartmental model for doxorubicin. These models were fitted to data using weighted least squares algorithms. The terminal phase rate constant kel was determined by the least-square method from the slope of terminal exponential decrease of drug concentration ( phase) to the time in semi-logarithmic coordinates. Area under the curve (AUC) of plasma concentrations was calculated using the trapezoidal method from the first to the last measurable concentration and extrapolated to infinity by dividing the last plasma concentration by the rate constant for elimination of the drug. To determine the influence of cinchonine on the pharmacokinetics of doxorubicin, each patient was considered as his own control and a Student's t-test for paired series was used. Statistical analysis of data obtained from Holter recording was performed using analysis of variance (ANOVA), according to the period alone or the period and the dose level of cinchonine. Differences between doses were analyzed using the Bonferini's test. Comparisons of parameters in terms of doses and as a function of the period were performed using Student's t-test for unpaired series. A calculated P value <0.05 was considered significant.

Results

Characteristics of patients

Thirty-four patients entered the study. It was initially planned to include three patients at each cinchonine level and to stop dose escalation when any grade 3 or 4 toxicity occurred in two out of three patients at a given dose level. Identification of increased QT intervals led us to increase the number of patients at each step better to monitor electric changes by a combination of repeated ECGs and the Holter technique (see below). The total number of patients included at each of the five different steps is indicated in Table 1. Each patient received at least one cycle of the cinchonine-CHVP combination. One to five additional cycles were administered every 4 weeks, based on the physician assessment of clinical benefit and potential toxicity and the patient's agreement. The total number of cycles administered at each of the five different steps is indicated in Table 1. Altogether, 87 cycles of the CHVP-cinchonine combination were administered. All the patients had been previously treated with anthracyclines and vinca alkaloids. The mean delay between initial diagnosis and inclusion in the study was 55 ± 44 (range 4-158) months. The main characteristics of these patients are summarized in Table 2.

Non-hematologic toxicity

All 34 patients and 87 cycles of treatment were available for toxicity analysis. Nineteen patients (56%) reported one or several transient symptoms previously related to quinine derivatives, including headache (nine patients, 19 cycles), dry mouth (nine patients, 12 cycles), tinnitus (two patients, three cycles) and dizziness (two patients, two cycles). These symptoms remained moderate, did not require specific treatment or cinchonine dose decrease and disappeared at the end of drug infusion. Fifteen of these 19 patients received a cinchonine dose equal to or higher than 25 mg/kg/d. Other adverse events were classified according to the NCI criteria and listed in Table 3. Abnormal liver function was registered in 15 (44%) patients during the first CHVP/cinchonine cycle and 43% of the 87 evaluated cycles. In all cases but three, liver toxicity remained limited to NCI grade 1-2 toxicity. A mild to moderate increase of creatininemia was identified in 23% of the cycles. In most cases, these patients had required nephrotoxic antibiotics and/or antifungal drugs in addition to chemotherapy. Infections were observed in 23% of the 87 cycles and occurred in the setting of disease-related cytopenia and immune deficiency. Nausea and vomiting were observed in 25 and 20% of the cycles respectively, despite the systematic administration of antiemetic drugs. In two patients, a decrease of the left ventricular ejection fraction was detected during the course of the study, after one and three cycles, respectively. One episode of grade 1 bradycardia and two episodes of supraventricular tachycardia were also registered. In other patients, ECG recording indicated that atrial rate (83 ± 12/min; range 52-103), PR interval (0.15 ± 0.03 s; range 0.1-0.2) and QRS duration (0.09 ± 0.01 s; range 0.08-0.10) were not significantly influenced by cinchonine administration at any of the tested doses. Cardiac monitoring detected a prolongation of the QT interval in the first patients included in the study. This event required a more specific analysis, the results of which are described below.

QT interval increase

Long-term ECG recording using the Holter technique was used to study QT interval further, to measure heart rate and to detect arrhythmia in 23 patients (total number of recordings: 136). The corrected QT interval (QTc), according to Bazett's formula,²⁹ significantly increased under cinchonine exposure, whether measured at the top of the T wave (QTac) (Figure 2) or at the end of the T wave (QTec) (data not shown). This increase did not correlate with age, sex, associated medications, previous dose of anthracycline received, potassium and magnesium serum level and previous history of any cardiac symptoms. QT interval increase appeared during the first hours of cinchonine infusion, remained stable during drug infusion and returned to the baseline level over a period of 8 to 10 h after discontinuation of the drug. However, at the 35 mg/kg/d cinchonine dose, QTc interval remained significantly increased during period D, as compared with period A (P < 0.05). Moreover, at this dose level, five of eight patients included demonstrated a more than 20% increase of the QT interval (Table 4). ANOVA indicated a significant correlation between the dose level of cinchonine and the QT interval increase (P = 0.013). No significant difference was observed between QT intervals registered during period B and those registered during period C, suggesting that doxorubicin did not influence QT duration (P = 0.768). Based on these observations, the dose escalation was stopped and 30 mg/kg/d was considered as the MTD of cinchonine administered by continuous i.v. infusion.

Other cardiac effects

The incidence of isolated supraventricular premature beats was reduced by a half during period B (0.97 ± 0.23/h) compared to period A (1.80 ± 0.55/h). The incidence of ventricular premature beats also decreased during cinchonine perfusion (0.35 ± 0.13 and 0.25 ± 0.13 VPBs/h during period B and C, respectively, vs 0.43 ± 0.19 VPBs/h during period A). Couplets of ventricular premature beats were registered in two patients. Long-term ECG recording also confirmed that cinchonine did not exhibit any significant bradycardiac effect.

Hematologic toxicity

A majority of the 34 studied patients demonstrated disease-related cytopenia at the time of inclusion in the trial (Table 5), including anemia (n = 28), granulocytopenia (n = 12) and thrombopenia (n = 26). The CHVP/cinchonine induced cytopenia in all patients with normal hemogram at inclusion and increased the severity of cytopenia in all other patients. In all cases the patients recovered from this transient decrease in peripheral blood cell count. Moreover, increased hemoglobin level and platelet count (as compared to their respective baseline level) were considered as a positive response criteria in 7 of 11 responding patients (see below).

Serious adverse events

A total of 34 serious adverse events (SAE) were reported in 22 out of 34 patients during the course of this study. The main cause for these SAE was fever (n = 21) that was related to documented infection in 14 cases (bacterial septicemia: six, pneumonia: three, cytomegalovirus infection: two, gastroenteritis: one, oral candidosis: one; invasive pulmonary aspergillosis: one). Fever was from unknown origin in the seven remaining cases. All these patients recovered under anti-infectious therapy. The origin of other SAE was supraventricular tachycardia (n = 1), cytopenia requiring transfusion (n = 4) and disease progression (n = 10) that led to death in six patients. None of the deaths that occurred during the trial could be related to cinchonine infusion.

Pharmacokinetics of cinchonine

The pharmacokinetics of cinchonine was assessed over a 69-h period, including the 48-h period of infusion and the 21 h following drug infusion disclosure (Table 6). This study was completed during the first course of cinchonine in 23 of the 34 patients included in the study. Cinchonine serum concentration dramatically increased during the first hours of drug infusion to reach a steady-state level after 8 to 12 h of drug infusion and decreased rapidly after the end of this infusion. Significant inter-individual variability was observed. Nevertheless, both the concentration of cinchonine at the steady-state level (Css: mean of 6 concentrations) and the area under the curve (AUC) increased with the total dose of cinchonine administered (Figure 3; r² = 0.388 and 0.387, respectively). Although no significant relationship was observed between the cinchonine dose expressed in mg/kg and the pharmacokinetic parameters, the higher mean values of Css and AUC were obtained at the higher cinchonine dose (Table 6). The concentration of the major metabolite detected on the HPLC profile could not be measured directly since the chemical synthesis of this metabolite has not yet been performed. In the absence of an available standard, we measured the ratio of the peak area of this metabolite to the peak area of cinchonine internal standard. This analysis suggested that the pharmacokinetic profile of the metabolite was similar to that of cinchonine (data not shown). Less than 10% of cinchonine administered was excreted in urines in which the major metabolite was the main compound identified.

MDR-reversing activity in serum samples

To assess the efficacy of cinchonine as an MDR-reversing agent in patients, we measured the MRA of every serum sample collected for the pharmacokinetic analysis.²⁴ A significant MRA was detected in the serum from every patient. This activity appeared during the first hours of drug infusion, reached a steady-state level after 8 to 12 h of cinchonine infusion and decreased rapidly after the end of drug administration (see for example Figure 4). Linear regression analysis demonstrated a good correlation between MRA and cinchonine serum level at all the doses tested (Table 7). The maximal MRA slightly increased with the dose of cinchonine administered, although this increase was not statistically significant (Table 7).

Pharmacokinetics of doxorubicin

The influence of cinchonine on the pharmacokinetics of doxorubicin was studied in six additional patients (five men, one woman, mean age; range 52-75 years) at the end of the present study. Each of these patients received one course of CHVP regimen without cinchonine. Three weeks later, they received a second course of the same chemotherapy in combination with a 20-h continuous i.v. infusion of 30 mg/kg/d cinchonine, starting 12 h before chemotherapy infusion. The pharmacokinetic profile of doxorubicin with and without cinchonine is shown in Figure 5 and the main pharmacokinetic parameters are summarized in Table 8. Cinchonine demonstrated limited influence on the pharmacokinetic parameters of doxorubicin since the only significant change was a 31% decrease of its total body clearance (P = 0.03).

Response to treatment

A clinical or biological improvement was observed in 11 of the 34 included patients and was associated in every case with an improved performance status (Table 9). Five patients with multiple myeloma demonstrated a significant decrease of their immunoglobulin serum level and/or an increase in their platelet count. Four patients with non-Hodgkin's lymphoma also demonstrated a clinical response characterized by a decrease of tumor volume and/or an improvement of biological markers such as LDH serum level. In two patients with anthracycline refractory CLL, we observed a decrease of their peripheral blood lymphocyte count and an increase of their hemoglobin level. Two additional transient responses were registered: one patient with CLL demonstrated a significant decrease in lymph node volume and spleen size without change in peripheral lymphocyte count whereas another patient with multiple myeloma showed a 50% fall of serum monoclonal immunoglobulin level after the first cycle of chemotherapy but further progressed after two cycles.

Discussion

The present study demonstrates that efficient serum concentrations of cinchonine, a cinchona bark alkaloid that modulates the function of P-glycoprotein in vitro, are achievable in patients when this compound is administered by continuous intravenous infusion. Cinchonine was observed to delay cardiac repolarization in a dose-dependent manner. This cardiac side-effect was considered as the dose-limiting toxicity.

P-glycoprotein inhibition was suggested to increase the cardiac toxicity of anthracyclines in animal models.^30,31,32 This observation was not confirmed in clinical trials of chemosensitizing agents.^33,34 In the present study, two of the 40 patients who received the CHVP/cinchonine regimen demonstrated a mild and asymptomatic decrease of their left ventricular ejection fraction during the course of the study and none demonstrated symptomatic heart failure. On the other hand, acute cardiac toxicity was frequently observed in studies testing the first chemosensitizing agents designated to modulate the function of P-glycoprotein.^35,36,37 For example, the calcium channel-blocking agent verapamil was observed to induce transient complete heart block, congestive heart failure and severe hypotension.³⁸ However, cardiovascular toxicity can be distinguished from P-glycoprotein inhibition, as demonstrated with verapamil chiral enantiomers^39,40 and several other MDR-reversing drugs.^33,34

Preclinical studies performed in rats and dogs did not anticipate the prolonged cardiac repolarization observed in most patients who received cinchonine.²⁶ Increased QTc intervals^29,41,42 were detected by systematic ECG monitoring. Class III antiarrhythmics such as amiodarone and bretylium intentionally increase cardiac repolarization⁴² whereas unintentional QTc increase has been related to syncope and potentially life-threatening ventricular tachycardia, especially 'torsades de pointes'.⁴³ Quinidine, which has been tested as an MDR-reversing agent,³⁷ is one of these medications that can provoke syncopes at low plasmatic concentrations.^44,45 The triazine-aminopiperidine derivative, S9788, that has been developed as a specific MDR-modulator, induced auriculo-ventricular blocks and 'torsades de pointes'.⁴⁶ It is worth noting that no ventricular arrhythmia was observed during the 96 cycles of cinchonine.

The antiarrhythmic potency varies among cinchona bark alkaloids.^47,48 Cinchonine exerted a discrete, anti-arrhythmic effect without causing bradycardia. A triplet of ventricular extrasystoles was detected in one patient whereas long-term ECG registration using the Holter method showed that the overall incidence of ventricular premature beats decreased during cinchonine perfusion. Frequential analysis indicated that control of the heart rate by the sympathetic nervous system was reduced by cinchonine, an effect observed with the class Ic anti-arrhythmic propafenone,⁴⁹ but did not significantly affect the sympathetico-vagal balance.^50,51 This may explain the absence of 'torsade de pointes' whose occurrence has been shown to require a low stimulation frequency.⁵² All the patients were systematically supplemented in potassium and magnesium, which could also account for the lack of arrhythmia.⁵³

A significant correlation was observed between the dose of cinchonine administered and the QTc interval increase. At the 35 mg/kg/d cinchonine dose, the mean QTc interval remained increased during the last period of observation. At this dose level, five of eight patients demonstrated a more than 20% increase of QTc interval. Based on these two observations, the dose escalation was stopped at 35 mg/kg/d and we considered that 30 mg/kg/d was the MTD.

Measurement of the MRA in serum from patients receiving an MDR reversing drug is an interesting indication of the drug activity in the clinical setting.^17,18,19 A 12-h cinchonine infusion was required to reach an optimal MRA. After cinchonine infusion arrest, both drug serum level and MRA decreased rapidly, suggesting that cinchonine infusion might be maintained for at least 3 h after chemotherapeutic drug administration. We also show that cinchonine only slightly decreases doxorubicin clearance, in contrast with the strong increase of doxorubicin AUC observed in the presence of dexverapamil.³⁹ Several other MDR reversing agents were shown to decrease the clearance of cytotoxic drugs.^{18,39,54,55,56,57,58,59} However, the pattern of alteration of cytotoxic pharmacokinetics varies among the P-glycoprotein inhibitors, suggesting inhibition of specific cytochrome P-450 drug metabolism or interaction with membrane efflux pumps other than P-glycoprotein rather than a direct consequence of P-glycoprotein inhibition.^60,61

Although MDR phenotype is a prognostic factor in several malignant diseases, the issue of whether P-glycoprotein is a useful therapeutic target or a marker of cancer cell behavior remains a controversial issue⁶² that requires potent MDR-reversing drugs to be evaluated. When compared to quinine, cinchonine has the advantage of inducing a more efficient P-glycoprotein inhibiting activity in patient serum. Compared to cyclosporin A and its analogue PSC833, cinchonine has the advantage of having limited influence on the pharmacokinetics of doxorubicin. These characteristics justify further development of the compound. A phase II study has been initiated by the GELA group in patients with refractory or relapsed non-Hodgkin's lymphomas. Given the potential cardiac toxicity of cinchonine, this phase II trial includes careful cardiac monitoring as well as magnesium and potassium supplementation to prevent any ventricular arrhythmia.

Acknowledgements

The authors wish to thank Marie-Christine Verbist and Marie-Claire Pinel for their helpful participation in monitoring the study. INSERM U517 is supported by grants from ARC (grant No. 5817), the departemental comittees (Côte d'Or, Nièvre, Saône et Loire) of the Ligue Nationale Contre le Cancer, the Conseil Régional de Bourgogne and the Commission de Recherche Clinique du CHU de Dijon.

1 Filipits M, Suchomel RW, Zöchbauer S, Malayeri R, Pirker R. Clinical relevance of drug resistance genes in malignant diseases. Leukemia 1996; 10: S10-S17, MEDLINE

2 Gottesman MM, Pastan I, Ambudkar SV. P-glycoprotein and multidrug resistance. Curr Opin Genet Dev 1996; 6: 610-617, MEDLINE

3 Thiebaut F, Tsuruo T, Hamada H, Gottesman MM, Pastan I, Willingham MC. Cellular localization of the multidrug-resistance gene product P-glycoprotein in normal human tissues. Proc Natl Acad Sci USA 1987; 84: 7735-7738, MEDLINE

4 Cordon-Cardo C, O'Brien JP, Casals D, Rittman-Grauer L, Biedler JL, Melamed MR, Bertino JR. Multidrug-resistance gene (P-glycoprotein) is expressed by endothelial cells at blood-brainbarrier sites. Proc Natl Acad Sci USA 1989; 86: 695-698, MEDLINE

5 Chaudhary PM, Roninson IB. Expression and activity of P-glycoprotein, a multidrug efflux pump, in human hematopoietic stem cells. Cell 1991; 66: 85-94, MEDLINE

6 Van Kalken CK, Broxterman HJ, Pinedo HM, Feller N, Dekker H, Lankelma J, Giaccone G. Cortisol is transported by the multidrug resistance gene P-glycoprotein. Br J Cancer 1993; 67: 284-289, MEDLINE

7 Niehans GA, Jaczsz W, Brunetto V, Perri RT, Gajl-Peczalska K, Wick MR, Tsuruo T, Bloomfield CD. Immunohistochemical identification of P-glycoprotein in previously untreated, diffuse large cell and immunoblastic lymphomas. Cancer Res 1992; 52: 3768-3775, MEDLINE

8 Dalton WS, Grogan TM, Meltzer PS, Sheper RJ, Durie BGM, Taylor CW, Miller RP, Salmon SE. Drug resistance in multiple myeloma and non-Hodgkin's lymphoma: detection of P-glycoprotein and potential circumvention by addition of verapamil to chemotherapy. J Clin Oncol, 1989; 7: 415-424,

9 Yuen AR, Sikic BI. Multidrug resistance in lymphomas. J Clin Oncol 1994; 12: 2453-2459, MEDLINE

10 Sandor V, Wilson W, Fojo T, Bates SE. The role of MDR-1 in refractory lymphoma. Leuk Lymphoma 1997; 28: 23-31, MEDLINE

11 Sonneveld P, Lokhorst HM, Vossebeld P. Drug resistance in multiple myeloma. Semin Hematol 1997; 34: S34-S39,

12 Wallner J, Gisslinger H, Gisslinger B, Gsur A, Götzl M, Zöchbauer S, Pirker R. MDR1 gene expression in chronic lymphocytic leukemia. Leuk Lymphoma 1994; 13: 333-338, MEDLINE

13 Epstein J, Xiao H, Oba BK. P-glycoprotein expression in plasma-cell myeloma is associated with resistance to VAD. Blood 1989; 74: 913-917, MEDLINE

14 Marie JP, Zittoun R, Sikic BI. Multidrug resistance (mdr1) gene expression in adult acute leukemias: correlations with treatment outcome and in vitro drug sensitivity. Blood 1991; 78: 586-592, MEDLINE

15 Campos L, Guyotat D, Archimbaud E, Calmard-Oriol P, Tsuruo T, Troncy J, Treille D, Fiere D. Clinical significance of multidrug resistance P-glycoprotein expression on acute nonlymphoblastic leukemia cells at diagnosis. Blood 1992; 79: 473-476, MEDLINE

16 Baldini N, Scotlandi K, Barbanti-Brodano G, Manara MC, Maurici D, Bacci G, Bertoni F, Picci P, Sottili S, Campanacci M, Serra M. Expression of P-glycoprotein in high-grade osteosarcomas in relation to clinical outcome. New Engl J Med 1995; 333: 1380-1385, MEDLINE

17 Ford JM, Hait WN. Pharmacology of drugs that alter multidrug resistance in cancer. Pharmacol Rev 1990; 42: 155-199, MEDLINE

18 Lum BL, Fisher GA, Brophy NA, Yahanda AM, Adler KM, Kaubisch S, Halsey J, Sikic BI. Clinical trials of modulation of multidrug resistance. Pharmacokinetic and pharmacodynamic considerations. Cancer 1993; 72: 3502-3514, MEDLINE

19 Chauffert B, Pelletier H , Corda C, Solary E, Bedenne L, Caillot D, Martin F. Potential usefulness of quinine for the circumvention of the anthracycline resistance in clinical practice. Br J Cancer 1990; 62: 395-397, MEDLINE

20 Solary E, Velay I, Chauffert B, Bidan JM, Caillot D, Dumas M, Guy H. Sufficient levels of quinine in the serum circumvent the multidrug resistance of the human leukemic cell line K562/ADM. Cancer 1991; 68: 1714-1719, MEDLINE

21 Solary E, Caillot D, Chauffert B, Casasnovas RO, Dumas M, Maynadie M, Guy H. Feasibility of using quinine, a potential multidrug resistance-reversing agent, in combination with mitoxantrone and cytarabine for the treatment of acute leukemia. J Clin Oncol 1992; 10: 1730-1736, MEDLINE

22 Solary E, Witz F, Caillot D, Moreau P, Desablens B, Cahn JY, Sadoun A, Berthou C, Maloisel F, Guyotat D, Casassus P, Ifrah N, Lamy T, B Audhuy B, Colombat P, Harousseau JL. Combination of quinine as a potential reversing agent with mitoxantrone and cytarabine for the treatment of acute leukemias: a randomised multicentric study. Blood 1996; 88: 1198-1205, MEDLINE

23 Wattel E, Solary E, Hecquet B, Caillot D, Ifrah N, Brion A, Mahé B, Milpied N, Janvier M, Guerci A, Rochant H, Cordonnier C, Dreyfus F, Buzyn A, Hoang-Ngoc L, Stoppa AM, Gratecos N, Sadoun A, Stamatoulas A, Tilly H, Brice P, Maloisel F, Lioure B, Desablens B, Pignon B, Abgrall JF, Leporrier M, Dupriez B, Guyotat D, Lepelley P, Fenaux P. Quinine improves the results of intensive chemotherapy in myelodysplatic syndromes expressing P-glycoprotein: results of a randomized study. Br J Haematol 1998; 102: 1015-1024, MEDLINE

24 Genne P, Dimanche-Boitrel MT, Mauvernay RY, Gutierrez G, Duchamp O, Petit JM, Martin F, Chauffert B. Cinchonine, a potent efflux inhibitor to circumvent anthracycline resistance in vitro and in vivo. Cancer Res 1992; 52: 2797-2801, MEDLINE

25 Genne P, Duchamp P, Solary E, Pinard D, Belon JP, Dimanche-Boitrel MT, Chauffert B. Comparative effects of quinine and cinchonine in reversing multidrug resistance on human leukemic cell line K562/ADM. Leukemia 1994; 8: 160-164, MEDLINE

26 Genne P, Duchamp P, Solary E, Magnette J, Belon JP, Chauffert B. Cinchonine per os: efficient circumvention of P-glycoprotein-mediated multidrug resistance. Anti-Cancer Drug Design 1995; 10: 103-118, MEDLINE

27 Taggart JV, Earle DP, Berliner RW, Zubrod G, Welch WJ, Bowman Wise N, Schroder EF, Lond IM, Shannon JA. Studies on the chemotherapy of human malaria. III. The physiological disposition and antimalarial activity of cinchona alkaloids. J Clin Invest 1948; 27: 80-86,

28 Abraham I, Ueda CT. Comparative disposition kinetics of two diastereoisomeric pairs of cinchona alkaloids in the dog. J Pharmac Sci 1984; 73: 452-456,

29 Bazett HC. An analysis of time-relations of electrocardiograms. Heart 1920; 7: 353,

30 Bright JM, Buss DD. Effects of verapamil on chronic doxorubicin-induced cardiotoxicity in dogs. J Natl Cancer Inst 1990; 82: 963-964, MEDLINE

31 Bellamy WT, Peng YM, Odeleye A, Ellsworth L, Xu MJ, Grogan TM, Weinstein RS. Cardiotoxicity in the SCID mouse following administration of doxorubicin and cyclosporin A. Anticancer Drugs 1995; 6: 736-743, MEDLINE

32 van Asperen J, van Tellingen O, Tijssen F, Schinkel AH, Beijnen JH. Increased accumulation of doxorubicin and doxorubicinol in cardiac tissue of mice lacking mdr1a P-glycoprotein. Br J Cancer 1999; 79: 108-113, Article MEDLINE

33 Yahanda AM, Alder KM, Fisher GA, Brophy NA, Hasley J, Hardy RI, Gosland MP, Lum BL, Sikic BI. Phase I trial of etoposide with cyclosporine as a modulator of multidrug resistance. J Clin Oncol 1992; 10: 1624-1634, MEDLINE

34 Advani R, Saba HI, Tallman MS, Rowe JM, Wiernik PH, Ramek J, Dugan K, Lum B, Villena J, Davis E, Paietta E, Litchman M, Sikic BI, Greenberg PL. Treatment of refractory and relased acute myelogenous leukemia with combination chemotherapy plus the multidrug resistance modulator PSC833 (Valspodar). Blood 1999; 93: 787-795, MEDLINE

35 Tsuruo T, Iida H, Tsukagoshi S, Sakurai Y. Overcoming of vincristine resistance in P388 leukemia in vivo and in vitro through enhanced cytotoxicity of vincristine and vinblastine by verapamil. Cancer Res 1981; 41: 1967-1972, MEDLINE

36 Chauffert B, Martin M, Hammann A, Michel MF, Martin F. Amiodarone-induced enhancement of doxorubicin and 4'-deoxydoxorubicin cytotoxicity to rat colon cancer cells in vitro and in vivo. Cancer Res 1986; 46: 825-830, MEDLINE

37 Wishart GC, Bissett D, Paul J, Jodrell D, Harnett A, Habeshaw T, Kerr DJ, Macham MA, Soukop M, Leonard RCF, Knepil J, Kaye SB. Quinidine as a resistance modulator of epirubicin in advanced breast cancer: mature results of a placebo-controlled randomized trial. J Clin Oncol 1994; 12: 1771-1777, MEDLINE

38 Pennock GD, Dalton WS, Roeske WR, Appleton CP, Mosley K, Plezia P, Miller TP, Salmon SE. Systemic toxic effects associated with high-dose verapamil infusion and chemotherapy administration. J Natl Cancer Inst 1991; 83: 105-110, MEDLINE

39 Wilson WH, Jamis-Dow C, Bryant G, Balis FM, Klecker RW, Bates SE, Chabner BA, Steinberg SM, Kohler DR, Wittes RE. Phase I and pharmacokinetic study of the multidrug resistance modulator dexverapamil with EPOCH chemotherapy. J Clin Oncol 1995; 13: 1985-1994, MEDLINE

40 Echizen H, Brecht T, Niedergesass S, Vogelgesang B, Eichelbaum M. The effect of dextro-, levo- and racemic verapamil on atrioventricular conduction in humans. Am Heart J 1985; 109: 210-217, MEDLINE

41 Garson A. How to measure the QT interval: what is normal? Am J Cardiol 1993; 72: 14B-16B, MEDLINE

42 Bonate PL, Russell T. Assessment of QTc prolongation for non-cardiac-related drugs from a drug development perspective. J Clin Pharmacol 1999; 39: 349-358, MEDLINE

43 Algra A, Tijssen JGP, Roelandt J, Pool J, Lubsen J. QTc prolongation measured by standard 12-lead ECG is an independent risk factor for sudden death due to cardiac arrest. Circulation 1991; 83: 1888-1894, MEDLINE

44 Horowitz LN, Greenspan AM, Spielman SR, Josephson ME. Torsades de pointes: electrophysiologic studies in patients without transient pharmacologic or metabolic abnormalities. Circulation 1981; 63: 1120-1128, MEDLINE

45 Jenzer HR, Hagemeijer F. Quinidine syncope: torsades de pointes with low quinidine plasma concentrations. Eur J Cardiol 1976; 4: 447-451, MEDLINE

46 Stupp R, Bauer J, Pagani O, Gerard B, Cerny B, Sessa C, Bastian G, Sarkany M, Schläpfer J, Giroux B, Leyvraz S. Ventricular arrhytmia and torsade de pointes: dose limiting toxicities of the MDR-modulator S9788 in a phase I trial. Ann Oncol 1998; 9: 1233-1242, MEDLINE

47 Small LD, Rosenberg H, Nwangu PU, Holcslaw TL, Stohs SJ. Partial synthesis of 6'hydroxycinchonine and its antiarrhythmic activity in mice. J Med Chem 1979; 22: 1014-1016, MEDLINE

48 Nwangwu PU, Holcslaw TL, Rosenberg H, Small LD, Stohs SJ. The antiarrhythmic activities of 6'hydroxycinchonine and 6'allyloxycinchonine compared with quinidine in mice. J Pharm Pharmacol 1979; 31: 488-489, MEDLINE

49 Lombardi F, Torzillo D, Sandrone G, Dalla Vecchia L, Finocchiaro ML, Bernasconi R, Capiello E. Beta-blocking effect of propafenone based on spectral analysis of heart variability. Am J Cardiol 1992; 70: 1028-1034, MEDLINE

50 Lown B, Verrier RL. Neural activity and ventricular fibrillation. New Engl J Med 1976; 295: 1165-1170, MEDLINE

51 Nolan J, Flapan AD, Goodfield NE, Prescott RJ, Bloomfield P, Neison JMM, Ewing DJ. Measurement of parasympathetic activity from 24-hour ambulatory elctrocardiograms and its reproductibility and sensitivity in normal subjects, patients with symptomatic myocardial ischemia, and patients with diabetes mellitus. Am J Cardiol 1996; 77: 154-158, MEDLINE

52 Coumel P, Rosengarten MD, Leclercq JF, Attuel P. Role of sympathetic nervous system in non-ischemic ventricular arrhythmias. Br Heart J 1982; 47: 137-147, MEDLINE

53 Fontaine G, Frank R, Grosgogeat Y. Torsades de pointes: definition and management. Mod Concepts Cardiovasc Dis 1982; 51: 103-108, MEDLINE

54 Lum BL, Kaubisch S, Yahanda AM, Adler KM, Jew L, Ehsan MN, Brophy NA, Halsey J, Gosland MP, Sikic BI. Alteration of etoposide pharmacokinetics and pharmacodynamics by cyclosporine in a phase I trial to modulate multidrug resistance. J Clin Oncol 1992; 10: 1635-1642, MEDLINE

55 Marie JP, Bastie JN, Coloma F, Faussat Suberville AM, Delmer A, Rio B, Delmas-Marsalet B, Leroux G, Casassus P, Baumelou E, Catalin J, Zittoun R. Cyclosporine A as a modifier agent in salvage treatment of acute leukemia (AL). Leukemia 1993; 7: 821-824, MEDLINE

56 Sonneveld P, Marie JP, Huisman C, Vekhoff A, Schoester M, Faussat AM, van Kapel J, Groenewegen A, Charnick S, Zittoun R, Lowenberg B. Reversal of multidrug resistance by SDZ PSC 833, combined with VAD (vincristine, doxorubicin, dexamethasone) in refractory multiple myeloma. A phase I study. Leukemia 1996; 10: 1741-1750, MEDLINE

57 Berg SL, Tolcher A, O'Shaughnessy JA, Denicoff AM, Noone M, Ognibene FP, Cowan KH, Balis FM. Effect of R-verapamil on the pharmacokinetics of paclitaxel in women with breast cancer. J Clin Oncol 1995; 13: 2039-2042, MEDLINE

58 El-Yazigi A, Berry J, Ezzat A, Wahab FA. Effect of tamoxifen on the pharmacokinetics of doxorubicin in patients with non-Hodgkin's lymphoma. Ther Drug Monit 1997; 19: 632-636, MEDLINE

59 Aebi S, Schnider TW, Los G, Heath DD, Darrah D, Kirmani S, McClay EF, D'Agostino H, Plaxe SC, Fink SC, De la Alas MM, Howell SB, Chisten RD. A phase II/pharmacokinetic trial of high-dose progesterone in combination with paclitaxel. Cancer Chemother Pharmacol 1999; 44: 259-265, MEDLINE

60 Harker WG, Slade DL, Dalton WS, Meltzer PS, Trent JM. Multidrug resistance in mitoxantrone-selected HL-60 leukemia cells in the absence of P-glycoprotein overexpression. Cancer Res 1989; 49: 4542-4549, MEDLINE

61 Lee K, Belinsky MG, Bell DW, Testa JR, Kruh GD. Isolation of MOAT-B, a widely expressed multidrug resistance-associated protein/canalicular multispecific organic anion transporter-related transporter. Cancer Res 1998; 58: 2741-2747, MEDLINE

62 Pinedo HM, Giaccone G. P-glycoprotein, a marker of cancer-cell behavior. New Engl J Med 1995; 333: 1417-1419, MEDLINE

Figure 1  Schematic representation of the trial. CHVP: cyclophosphamide, doxorubicin, vincristine, prednisone.

Figure 2  Time course of mean QTac. QTac was measured from the onset of the QRS to the top of the T wave and corrected according to Bazett's formula. The experimental periods named A, B, C and D were compared (see text).

Figure 3  Relationship between total dose of cinchonine administered and cinchonine concentration at the steady-state level calculated as the mean of 6 concentrations (Css, panel a, r² = 0.388) or area under the serum concentration-time curve (AUC, panel b, r² = 0.387).

Figure 4  Time course of cinchonine serum level (panel a) and multidrug resistance reversing activity (panel b) in one representative patient.

Figure 5  Influence of cinchonine on doxorubicin pharmacokinetics. Six patients received one course of CHVP regimen without cinchonine. Three weeks later, each received a second course of the same chemotherapy in combination with a 20 h continuous i.v. infusion of 30 mg/kg/d cinchonine, starting 12 h before chemotherapy infusion.

Table 1  Cinchonine dose escalation. Cinchonine was given by continuous i.v. infusion for 48 h and escalated over five dose levels

Table 2  Characteristics of patients

Table 3  Non-hematologic toxicity registered during the 87 cycles of the CHVP/cinchonine regimen

Table 4  Dose-dependent influence of cinchonine of QT interval corrected according to Bazett's formula (QTc)

Table 5  Hematological toxicity

Table 6  Pharmacokinetic parameters of cinchonine

Table 7  Pharmacodynamic parameters of cinchonine

Table 8  Influence of cinchonine on the pharmacokinetic parameters of doxorubicin

Table 9  Response to treatment