Kinetic targeting of pegylated liposomal doxorubicin: a new approach to reduce toxicity during chemotherapy (CARL-trial)

Abstract

Background: The therapeutic success of chemotherapeutic agents is often limited by severe adverse effects. To reduce toxicity of these drugs, nanoscale particle-based drug delivery systems (DDS) are used. DDS accumulate to some extent in tumor tissues, but only a very small portion of a given dose reaches this target. Accumulation of DDS in tumor tissues is supposed to be much faster than in certain other tissues in which side effects occur ("Kinetic Targeting"). Once saturation in tumor tissue is achieved, most of the administered DDS still circulate in the plasma. The extracorporeal elimination of these circulating nanoparticles would probably reduce toxicity.

Methods: For the CARL-trial (Controlled Application and Removal of Liposomal chemotherapeutics), pegylated liposomal doxorubicin (PLD) was used as chemotherapeutic agent and double filtration plasmapheresis (DFPP) was performed for extracorporeal elimination of liposomes. PLD was given as 40 mg/m2 every 3 weeks in combination with vinorelbine 2 × 25 mg/m2 (neoadjuvant treatment of breast cancer, 12 patients), or as 40 mg/m2 every 4 weeks (recurrent ovarian cancer, 3 patients). Primary endpoints were the efficiency and safety profile of DFPP, and secondary endpoints were side effects and tumor response.

Results: DFPP eliminated ~62% of circulating PLD, corresponding to ~45% of the total dose (n = 57 cycles). AUC of doxorubicin was reduced by 50%. No leakage of doxorubicin was detected during elimination, and no relevant DFPP-related side effects occurred. Reduction in tumor size > 30% occurred in 10/12 (neoadjuvant) and in 1/3 patients (recurrent). Only five grade 2 events and one grade 3 event (mucositis, neutropenia or leucopenia) and a single palmar-plantar erythrodysesthesia grade 2 were reported.

Conclusion: Extracorporeal elimination of PLD by DFPP is safe and efficient. CARL can diminish the main dose-limiting side effects of PLD, and probably many different DDS alike.

Trial registration: DRKS00000163.

Figure 1

Apheresis of liposomal doxorubicin. Chart A shows the concentration of doxorubicin during plasmapheresis in blood (red line), in the separated plasma prior to particle filtration (blue line) and post-particle filtration (black line). The insert shows a schema of the double filtration plasmapheresis used. Blood circuit is red and plasma circuit is yellow. First filtration separates plasma from blood cells, second filtration step eliminates particles. The plasmapheresis system returns the plasma and blood remaining in the extracorporeal unit to the patient once treatment is finished, and final blood concentration of doxorubicin was measured thereafter. Average blood flow rates were between 50 and 65 ml/min. A typical plasmapheresis is shown (n = 57). Chart B shows the estimated area under the plasma concentration curve (AUC) for liposomal doxorubicin without plasmapheresis (below dashed line) and with plasmapheresis (gray).

Figure 2

Major toxicities observed during the CARL trial and related published trials. The major levels of toxicity for different treatment schedules using a combination of PLD and vinorelbine are given as percentage of patients and as number of patients/total patients in trial. Grading according to CTCAEv3-criteria is given in Roman numerals. Black bars: study of Martin et al. (n = 34), vinorelbine 30 mg/m² and PLD 35 mg/m² every 4 weeks [20]; red bars: study of Burstein et al. (n = 21), vinorelbine 2 × 25-30 mg/m² and PLD 40-50 mg/m² every 4 weeks [19]; blue bars: CARL-Trial (n = 12), vinorelbine 2 × 25 mg/m² and PLD 40 mg/m² + plasmapheresis every 3 weeks. Unfortunately, the frequency of neutropenia < grade IV is not reported in the study of Martin et al.. PPE = palmar-plantar erythrodysesthesia, PLD: pegylated liposomal doxorubicin.

Figure 3

QLQ-C30 Scores. To address quality of life during chemotherapy, the QLQ-C30 questionnaire was used [17]. The scores were calculated according to the manual. The questionnaire was filled after the first cycle (timepoint 1) and after the last cycle (timepoint 2), 10 patients were eligible completing 4 (n = 9) or 6 (n = 1) cycles. (A) Functional scores (high score corresponds to high functionality) (B) Symptom scores (low score correspond to mild symptoms).