Phase II Study on Bortezomib (BTZ) in Multiple Relapsed or Refractory Pediatric Acute Lymphoblastic Leukemia (rALL): High Response Rate with a Modestly Intensive Regimen Including BTZ, Not Related to Pharmacokinetics

BTZ is a proteasome inhibitor and could improve the poor prognosis of rALL (Messinger 2012). We developed a European phase II study in pediatric rALL, in which patients were randomised for ‘early’ or ‘late’ BTZ, in compliance with the Declaration of Helsinki. BTZ was given as iv push for 4 doses at 1.3 mg/m²/dose, in group ‘early’ on days 1, 4, 8 and 11 and in group ‘late’ on days 8, 11, 15 and 18. All patients also received DXM (10 mg/m2/day in 3 doses for 2 weeks) and VCR (1.5 mg/m2/dose (maximum 2 mg) as 1-hour infusion on days 8 and 15), plus intrathecal methotrexate (i.t. MTX, dose age-adjusted) on day 1. Cycles could be repeated in case of a good response. The number of 12 patients in each group provides a power of 80% to detect a statistically significantly lower absolute number of circulating leukemic blasts on day 8 of treatment in patients exposed to ‘BTZ early’ as compared to patients not yet exposed to BTZ (‘BTZ late’) of at least 30% (2-sided test, alpha <0.05).

Between October 2010 and September 2014, 29 patients with information on response and/or toxicity after cycle 1 enrolled, 17 boys and 12 girls, median 9.7 years of age (range, 1 – 17.5). Most had second or subsequent rALL (n=16). The primary endpoint, day 8 absolute blast count, was evaluable in 26 patients (n=2 not eligible for the primary endpoint in retrospect, n=1 no material received), and was not different between ‘BTZ early’ (n=13) and ‘BTZ late’ (n=13). Absolute blast count had decreased at day 8 in both ‘BTZ early’ (p=0.017) and ‘BTZ late’ (p=0.046) compared to day 1 blast count, but to the same extent (median 85-86% decrease).

Regarding response after cycle 1, 25 patients were evaluable (n=2 discontinued treatment due to toxicity, n=2 no material received). Nine patients (36%) achieved a complete remission with incomplete blood count recovery (CRi), 6 (24%) a partial remission (PR), and 10 had a treatment failure (TF), for a total response rate of 60%. There was no relation between remission status and the administration of BTZ being early or late (p=0.7). There was no association between response and disease status, although numbers were small (p=0.8). Four patients with a relapse after allo-SCT achieved CRi, as well as 5 patients with a second or subsequent relapse.

Grade 3 or 4 hematological toxicities were thrombopenia (n=29), leukopenia (n=23), neutropenia (n=25), and anemia (n=21). Most common grade 2-4 neurotoxicities were pain (n=9) and peripheral neuropathy (PN) (n=5). Thirteen patients received a second cycle, when one patient developed grade 2 PN and one patient with PN advanced from grade 2 to 3.

There was remarkable intra- and inter-individual variability in peak plasma concentrations 15 min after the first administration of BTZ, ranging from 0.7 to 715 ug/L (median 12 ug/L, excluding 2 outliers above limit of quantification (2500 ug/L)). The area-under the plasma concentration curve (AUC) had a 25-fold interindividual variation: median 0.27 ug.h/L (range 0.07 – 1.77). Mean clearance in plasma was 1.1 L/h, with a interindividual variability of 95%. Median bone marrow concentrations of BTZ were 49 ug/kg (range: 1.6 – 106 ug/kg). In most patients, no BTZ was detected in the CSF, with a lower detection limit of 0.05 ng/ml. In 9 patients BTZ was detected at 0.07 – 5.2 ug/L. No correlation between AUC and concentration in CSF was found (r2 = 0.11). The 50% inhibitory concentration (IC₅₀) of 20S activity in plasma was 10.7 ug/L. Plasma pharmacokinetic and pharmacodynamic parameters did not correlate with response to treatment.

In conclusion, BTZ in combination with VCR and DXM and i.t. MTX is effective in a significant subset of pediatric rALL, with 60% of patients achieving PR or CRi. Because administration of BTZ early or late showed similar early treatment responses, randomized studies with more patients are necessary to determine the additive value of BTZ to conventional chemotherapy. Repetitive cycles could be given in several children without undue toxicity. In view of the observed toxicity, it might be wise to give these patients subcutaneous BTZ or novel proteasome inhibitors. BTZ does not or hardly penetrate the cerebrospinal fluid.

This research is financially supported by the Dutch Foundation Children Cancer-free (clinical research support) and by Janssen Pharmaceuticals (clinical research support and free drug).