Paclitaxel synergizes with exposure time adjusted CD22-targeting immunotoxins against B-cell malignancies

Abstract

CD22-targeted recombinant immunotoxins (rIT) are active in hairy cell leukemia or acute lymphoblastic leukemia (ALL), but not in mantle cell lymphoma (MCL) patients. The goal was to enhance rIT efficacy in vivo and to define a strong combination treatment. Activity of Moxetumomab pasudotox (Moxe) and LR combined with paclitaxel was tested against MCL cell lines in vitro and as bolus doses or continuous infusion in xenograft models. In the KOPN-8 ALL xenograft, Moxe or paclitaxel alone was active, but all mice died from leukemia; when combined, 60% of the mice achieved a sustained complete remission. Against MCL cells in vitro, LR was more active than Moxe and the cells had to be exposed to rIT for more than 24 hours for them to die. To maintain high blood levels in vivo, LR was administered continuously by 7-day pumps achieving a well-tolerated steady plasma concentration of 45 ng/ml. In JeKo-1 xenografts, continuously administered LR was 14-fold more active than bolus doses and the combination with paclitaxel additionally improved responses by 135-fold. Maintaining high rIT-plasma levels greatly improves responses in the JeKo-1 model and paclitaxel substantially enhances bolus and continuously infused rIT, supporting a clinical evaluation against B-cell malignancies.

Keywords: CD22-targeted immunotoxin; combination therapy; mantle cell lymphoma; paclitaxel; targeted therapy.

Figure 1. Paclitaxel enhances

in vivo efficacy of CD22-targeting immunotoxin in KOPN-8 xenografts. NSG mice bearing luciferase positive KOPN-8 cells were treated with a single dose of 25 mg/kg paclitaxel i.p. on day 7, 3 doses of 0.4 mg/kg Moxe QOD from day 8, or the combination. A. Bioluminescence signal was measured at the indicated days. All images were taken with identical camera settings (binning “medium”, exposure time 1 second). Signal intensity is shown as radiance from 100 (blue) to 30,000 (red). Mice without any detectable bioluminescence were imaged a second time using “high sensitivity” camera settings (binning “high”, exposure time 1 minute). The arrow points towards the “high sensitivity” image of mouse one of the Moxe group. B. Animal survival is shown as a Kaplan Meyer curve, the significance was determined by log-rank test.

Figure 2. The time for rITs to reach maximal cytotoxic activity is highly variable

Indicated MCL cell lines were exposed to 140 ng/ml Moxe (black) or equimolar LR (grey) for various times, washed, replated, and cell-viability determined after a total of 3 days. Viability of untreated cells was set to 100%, 0% was defined as real 0. Each bar represents the average percent living cells of three independent experiments, error is shown as SEM.

Figure 3. LR in citrate buffer is stable, pump-administration achieves high serum levels increasing activity by 10-fold

LR at 2 mg/ml in standard buffer PBS A. or citrate buffer B. were incubated at 37°C for the indicated times, frozen at -80°C, and activity of each sample determined by WST8 assays. Activity of an immediately frozen aliquot (day 0) was set to 1 to which the remainder were normalized. Bars indicate average normalized IC₅₀, errors are shown as SEM. C. Mice were implanted with an ALZET osmotic pump containing LR at 1 mg/ml in citrate buffer on day 0. From day 1, blood was collected, the plasma separated by centrifugation, and the LR concentration was determined by WST8 assays. Each symbol represents the average LR plasma concentration of 3 mice, except day 6 (only 2 mice), error is shown as SEM (except day 6), AUC as determined by Graph Pad Prism, v.6.01. The average plasma concentration is indicated as a dashed line. D. Some mice bearing systemic Jeko-1 were euthanized on day 0 and the JeKo-1 BM infiltration determined. The remainder were either implanted with a 7-day ALZET osmotic pump containing 1 mg/ml LR in citrate buffer and a pump rate of 0.5 μl/h, treated with 3 doses of 2.0 mg/kg LR QOD i.v. from day 0, or treated with vehicle. The treated mice were sacrificed on day 9 and their BM was analyzed for JeKo-1 cells by flow cytometry. Symbols indicate individual mice, lines represent mean JeKo-1 BM-infiltration, error as SEM. Significance determined by unpaired t-test as ns (not significant) or as indicated.

Figure 4. Minor

in vitro enhancement contrasts with a substantial improvement of in vivo efficacy by the addition of paclitaxel. A. MCL cell lines were treated with LR and various concentrations of paclitaxel; after 72 hours, cells were stained with Annexin V-PE and 7-AAD and viability was determined by flow cytometry. The bars represent the mean IC₅₀-fold changes when paclitaxel was added. Fold-change values were only included if paclitaxel alone reduced cell viability by at least 15% (indicating that paclitaxel at this concentration was active) but not more than 85% (non-linear regression becomes unreliable below this threshold). The fold-changes were summarized from at least three independent experiments; errors as SEM, p-values were determined by unpaired t-tests. B. Some JeKo-1 bearing mice were analyzed for BM infiltration at treatment start on day 0 and the remainder treated with either vehicle, a single dose of 25 mg/kg paclitaxel i.p. on day 2, implanted with a 7-day ALZET osmotic pump containing 1 mg/ml LR in citrate buffer from day 0, or the combination of pump and paclitaxel. Symbols indicate individual mice, lines represent mean JeKo-1 BM-infiltration, errors as SEM, p-values were determined by unpaired t-test. C. The graph summarizes 95 mice from a total of 6 individual experiments. Each bar represents one mouse and its fold-change of JeKo-1 BM infiltration 9 days after the indicated treatment relative to the average BM-infiltration at the treatment start of the corresponding experiment control on day 0. The color indicates treatment type as vehicle (black), Taxol mono (blue), 2.0 mg/kg LR i.v. QOD (yellow), LR 0.6 mg/kg/day by 7-day pump (green), and the combination of LR by pump and paclitaxel (red). D.-G. To determine the relative efficacy, the average BM-infiltration in the indicated groups was set to 1 and the relative fold-change was determined as (D) vehicle treated on day 9 (D9) vs. paclitaxel (D9), as (E) day 0 vs. QOD (D9), (F) as QOD (D9) vs. pump-administrated LR (D9), and as (G) pump-administrated LR without vs. with paclitaxel (D9). Color codes in (D-G) are as indicated in (C), the bars show the mean relative fold-change, errors are shown as SEM, p-values were determined by unpaired t-tests.