Exploring drug delivery for the DOT1L inhibitor pinometostat (EPZ-5676): Subcutaneous administration as an alternative to continuous IV infusion, in the pursuit of an epigenetic target
Nigel J. Waters a, Scott R. Daigle a, Bruce N. Rehlaender b, Aravind Basavapathruni a, Carly T. Campbell a,
Tyler B. Jensen a, Brett F. Truitt a, Edward J. Olhava c, Roy M. Pollock d, Kim A. Stickland a,⁎, Angelos Dovletoglou a
aEpizyme, Inc., 400 Technology Square, Cambridge, MA 02139, United States
bPharmadirections, 5001 Weston Parkway, Suite 103, Cary, NC 27513, United States
cCurrent address: Third Rock Ventures, 29 Newbury St. #3, Boston, MA 02116, United States
dCurrent address: Warp Drive Bio, LLC, 400 Technology Square, Cambridge, MA 02139, United States
a r t i c l e i n f o a b s t r a c t
Article history: Received 30 July 2015
Accepted 12 September 2015 Available online xxxx
Chemical compounds studied in this article: Pinometostat (EPZ-5676)
Keywords: EPZ-5676 Pinometostat MLL-r
DOT1L Subcutaneous Continuous IV
Protein methyltransferases are emerging as promising drug targets for therapeutic intervention in human can- cers. Pinometostat (EPZ-5676) is a small molecule inhibitor of the DOT1L enzyme, a histone methyltransferase that methylates lysine 79 of histone H3. DOT1L activity is dysregulated in the pathophysiology of rearranged mixed lineage leukemia (MLL-r). Pinometostat is currently in Phase 1 clinical trials in relapsed refractory acute leukemia patients and is administered as a continuous IV infusion (CIV). The studies herein investigated alterna- tives to CIV administration of pinometostat to improve patient convenience. Various sustained release technolo- gies were considered, and based on the required dose size as well as practical considerations, subcutaneous (SC) bolus administration of a solution formulation was selected for further evaluation in preclinical studies. SC ad- ministration offered improved exposure and complete bioavailability of pinometostat relative to CIV and oral ad- ministration. These findings warranted further evaluation in rat xenograft models of MLL-r leukemia. SC dosing in xenograft models demonstrated inhibition of MLL-r tumor growth and inhibition of pharmacodynamic markers of DOT1L activity. However, a dosing frequency of thrice daily (t.i.d) was required in these studies to elic- it optimal inhibition of DOT1L target genes and tumor growth inhibition. Development of an extended release formulation may prove useful in the further optimization of the SC delivery of pinometostat, moving towards a more convenient dosing paradigm for patients.
© 2015 Elsevier B.V. All rights reserved.
1.Introduction
Rearranged mixed lineage leukemia (MLL-r) involves recurrent translocations of the 11q23 locus and leads to an aggressive form of acute leukemia with limited chemotherapeutic options and a poor prognosis [1]. 11q23 translocations target the MLL gene and result in an oncogenic fusion protein comprising the amino-terminus of MLL fused in frame to any of over 70 potential fusion partners [2–4]. The MLL gene encodes a large multidomain protein with methyltransferase activity specific for lysine 4 on histone H3 (H3K4). Methylation of H3K4 regulates transcriptional activity of a characteristic set of genes includ- ing HOXA9 and MEIS1 [5–7]. Chromosomal rearrangements of the MLL gene result in the loss of MLL methyltransferase activity and fusion with members of the AF and ENL family of proteins. These fusion prod- ucts either directly or indirectly recruit DOT1L into complexes that pro- mote transcription elongation [8–10]. DOT1L catalyzes the methylation
of lysine 79 on histone H3 (H3K79), a chromatin modification that acti- vates gene transcription [11,12]. Suppression of DOT1L activity through genetic knockdown or small molecule inhibition leads to selective growth inhibition of MLL-r leukemia cells with minimal proliferative ef- fect on non MLL-r cell lines [13–20].
Pinometostat is an aminonucleoside modulator of DOT1L histone methyltransferase activity with picomolar potency (enzyme inhibition) and has been shown to reverse disease progression in animal models of MLL-r [13]. Pharmacokinetic characterization of pinometostat demon- strated low oral bioavailability and relatively high clearance in mouse and rat [13,21]. Previously, in vivo efficacy studies using intraperitoneal (IP) dosing had revealed that continuous maintenance of plasma expo- sure above a threshold level was likely to be required for efficacy. Ad- ministering EPZ-5676 as an IV infusion in a nude rat xenograft model of MLL-rearranged leukemia led to complete and sustained tumor re- gressions at doses that were well tolerated with no overt signs of toxic- ity or significant body weight loss. In the highest dose group, complete
⁎ Corresponding author.
E-mail address: [email protected] (K.A. Stickland).
http://dx.doi.org/10.1016/j.jconrel.2015.09.023 0168-3659/© 2015 Elsevier B.V. All rights reserved.
regressions were achieved following 21 days of dosing and tumor regrowth was not observed for the duration of the experiment.
Shortening the length of dosing to 14 days also caused tumor regres- sion; however, further reducing the duration to 7 days did not show sig- nificant efficacy. In addition, we found that intermittent dosing for 8 h
daily was less efficacious than a similar daily dose administered contin- uously over a 24-hour period. Therefore, our results suggested that con- tinuous inhibition of DOT1L activity for at least 14 days is required for
Fig. 1. Comparison of sustained release options and subcutaneous approaches for pinometostat. (A) Comparison of route of administration vs bioavailability, capacity (dose) and control of release for pinometostat; (B) subcutaneous approaches for pinometostat using infusion pumps, implants, micro-particles and nano-particles.
optimal effi cacy [13]. Based on these results, pinometostat was ad- vanced into a phase 1 clinical trial in relapsed refractory acute leukemia patients, administered as a 21- or 28-day continuous intravenous infu- sion. Compact portable infusion pumps make this regimen feasible even for ambulatory patients, but it is far from ideal. This report discusses efforts undertaken thus far to develop a more patient- convenient format.
Given the lack of oral bioavailability, it was highly unlikely that any route of delivery that required crossing an epithelium (oral, nasal, buc- cal, etc.) would be a viable option (Fig. 1A). Likewise, pulmonary deliv- ery could be ruled out due to the requirement for very low doses by this route. Hence, the technologies that remained as potentially viable op- tions were those that involved sustained release of drug in the blood- stream or into an extravascular space. Numerous technologies have been used to provide sustained release of active compounds following subcutaneous or intramuscular injection. These include PLGA [(poly(lactic-co-glycolic acid)] microspheres, other polymer micro particles, in situ gels, and lipid-based multiparticulate systems. While most of these are excellent for delivering relatively potent compounds over the course of a month or longer, none of them ap- peared feasible for a dose in the hundreds of milligrams per day range, as was required in this case. Various types of implants could also be considered, but the dosage of pinometostat required would require frequent surgical implantation (Fig. 1B). Given these consid- erations, the best option appeared to be a daily injection of a formu- lation containing a high drug load and providing more rapid release than other technologies.
2.Methods
2.1.Formulation HPLC analysis
Pinometostat and its related substances were analyzed using a re- verse phase gradient HPLC method with detection at 254 nm. The meth- od utilized high pH in the mobile phases to facilitate separation of pinometostat from its diastereomer. It was discovered during method development that in its free base form, pinometostat can chelate ferrous ion leaching from stainless steel components of the HPLC system; hence, EDTA was added to the mobile phase to address this issue. The analytical method was used for both the drug substance and the drug product and has been fully validated.
2.2.Solubility studies
Solubility studies were conducted using a low volume adaptation of the classical shake test method. Aliquots of pinometostat were placed in separate 2 mL vials, and 1 mL of 200 mM citrate buffer, with or without other excipients was added and the vial sealed. Vials were mechanically shaken for 24 h at room temperature, after which time samples were re- moved and filtered through a 0.45 μm PVDF filter. Filtrates were then di- luted and analyzed by HPLC. The final pH of the filtrate was determined using a micro-tip pH probe.
2.3.In vivo pharmacokinetics
All animal studies were conducted as per approved IACUC protocols.
2.4.Pharmacokinetic study in rat
The pharmacokinetics of pinometostat were evaluated in male Sprague–Dawley rats (n = 3 per dose route, 250–270 g, purchased from SLAC Laboratory Animal Co. Ltd). For subcutaneous administra- tion, pinometostat drug product, prepared in saline, was administered by subcutaneous injection. Serial blood sampling was employed in each animal at each time-point, 0.05, 0.22, 0.50, 1, 2, 4, 8 and 24h fol- lowing administration, with 150 μL of blood collected via the tail vein
into pre-chilled K2-EDTA tubes. Blood samples were put on wet ice and centrifuged at 4 °C (2000 × g for 5 min) to obtain plasma within 15 min of sample collection. Plasma samples were stored at – 80 °C prior to LC-MS/MS analysis.
2.5.Pharmacokinetic study in dog
The pharmacokinetics of pinometostat was evaluated in male beagle dogs (n = 3 per dose route, 6.9–9.0 kg, purchased from Beijing Marshall Biotechnology Co. Ltd). For subcutaneous administration, pinometostat drug product prepared in saline was administered by subcutaneous in- jection. Serial blood sampling was employed in each animal at each time-point, 0.083, 0.25, 0.5, 1, 2, 4, 8 and 24h following administration, with 500 μL of blood collected via the marginal vein into pre-chilled K2-EDTA tubes. Blood samples were put on wet ice and centrifuged at 4 °C (2000 × g for 5 min) to obtain plasma within 30 min of sample collection. Plasma samples were stored at – 80 °C prior to LC–MS/MS analysis.
2.6.LC–MS/MS bioanalysis and pharmacokinetic data analysis Pinometostat was extracted from K2-EDTA plasma by protein pre-
cipitation, using an acetonitrile-containing internal standard (dexa- methasone at a concentration of 50 ng/mL). Typically, samples were injected onto an LC-MS/MS system using an Acquity UPLC BEH C18 col- umn. The aqueous mobile phase was water with 0.025% formic acid and 1 mM ammonium acetate (A), and the organic mobile phase was meth- anol with 0.025% formic acid and 1 mM ammonium acetate (B). The gradient was as follows: 10% B for the first 0.2 min, increased to 98% B from 0.2 to 0.6 min, maintained at 98% B for 0.6 min, and decreased to 10% B within 0.01 min. The injection volume was 2 μL, and the total run time was 1.8 min with a fl ow rate of 0.6 mL/min. The retention time of pinometostat was 0.93 min. The ionization was conducted in the positive ion mode using the multiple reaction monitoring (MRM) transition [M + H]+ m/z 563.3 parent ion to m/z 326.3 daughter ion. Ten calibration standards were prepared in blank plasma of the relevant species providing a typical standard curve concentration range of 0.5–1000 ng/mL. Calibration curves were performed in duplicate in each analytical run together with low, mid and high concentration QCs in duplicate. All standard and QC measured concentrations fell within 85–115% of the nominal concentration. Pharmacokinetic param- eters were calculated by non-compartmental methods using WinNonlin (version 5.3; Pharsight, St. Louis, Missouri). Half-life (t1/2) values were determined by regression of at least three data-points in the later phase of the time–concentration profile. Parameters are presented as mean ± standard deviation where applicable.
2.7.Mouse MV4-11 xenograft study
In vivo studies were conducted after review by the proper animal care and use committee at Charles River Discovery Research Services (Durham, NC). MV4-11 cells harboring a MLL-AF4 rearrangement were implanted subcutaneously into the right fl ank of female Fox ChaseSCID® mice(CB17/Icr-Prkdcscid/IcrIcoCrl,CharlesRiver).Pinometostat was delivered by subcutaneous injection into the left flank of the test animal, contra-laterally to the tumor implantation site. Animals were sepa- rated into either an efficacy or PK/PD cohort. Both cohorts were dosed sub- cutaneously with either a twice daily (b.i.d.) or t.i.d. schedule with 60 and 40 mg/kg respectively. The twice daily efficacy dose group was dosed for a single treatment at 80 mg/kg before switching to 60 mg/kg. The mice did not tolerate the higher dose well, and animals were switched to a lower dose. A control group received subcutaneous injections t.i.d. of the vehicle, 5% hydroxypropyl-β-cyclodextrin (HPBCD) in saline. Ef- fi cacy was determined after 21 days of drug treatment followed by a 7 day drug holiday. Animals assessed for PK/PD were dosed for 14 days and euthanized three hours after the last dose. All mice
were weighed and tumors calipered twice weekly until the end of study. At the completion of the study animals were euthanized by terminal cardiac puncture under carbon dioxide anesthesia. Eutha- nized mice were sampled for tumor tissue. Tumors were collected in an RNAase-free environment, bisected, snap frozen in liquid nitro- gen, pulverized and fi nally stored at – 80 °C.
2.8.Histone extraction
For isolation of histones from ground tumor powder, approximately 20 mg of tumor powder was lysed in 500 μL nuclear extraction buffer (10 mM Tris–HCl, 10 mM MgCl2, 25 mM KCl, 1% Triton X-100, 8.6% su- crose, plus a Roche protease inhibitor tablet 1836145). A 5 mm steel bead was added to each sample (Qiagen, 69989) and the samples were lysed on the TissueLyser (Qiagen, 85210) for 30 s at 30/s frequen- cy. Sample blocks were rotated 180° and lysed for another 30 s at 30/s frequency. The samples were incubated on ice for 5 min, and then nuclei were collected by centrifugation at 600 g for 5 min at 4 °C and washed once in ice cold phosphate buffered saline (PBS). Supernatant was re- moved and histones were extracted for one hour with 0.4 N ice cold sul- furic acid. Extracts were clarifi ed by centrifugation at 10,000 g for 10 min at 4 °C and transferred to a fresh micro centrifuge tube contain- ing 10× volume ice cold acetone. Histones were precipitated at – 20 °C for 2 h, pelleted by centrifugation at 10,000 g for 10 min, and resuspend- ed in 150 μL water. Histones were quantifi ed using the BCA protein assay (Pierce, 23225).
2.9.ELISA analysis on the inhibition of H3K79me2
ELISA was performed as previously described [13]. Briefly, histones were run in matched H3K79me2 and total H3 ELISAs to calculate total
levels of H3K79 methylation and total histone H3, respectively. The op- tical density of the H3K79me2 ELISA was normalized to that of the total H3 ELISA for each sample.
2.10.Quantitative real-time PCR
For the isolation of RNA, 10 mg of tumor powder was lysed in 600 μL RLT Lysis Buffer (Qiagen) and homogenized with a Qiashredder column (Qiagen, 79656) following the manufacturer’s protocol. Flow-through was collected and total RNA was isolated using the RNeasy Total RNA isolation kit (Qiagen, 74,106) according to manufacturer’s instructions. HOXA9 and MEIS1 mRNA levels were assessed and normalized to beta- 2-microglobulin by qRT-PCR as previously described [14].
3.Results
3.1.Formulation development
Pinometostat (see structure in Fig. 2A) is a basic compound with corresponding pKa values of about 3.8, 6.0 and 8.8. It is insoluble at neutral pH, but early studies showed some solubility increase in hydroxypropyl-beta-cyclodextrin (HPBCD) solutions. The initial tar- get product profi le, based on the need to provide doses of up to 1 g per day, was a 10-mL vial of 100 mg/mL formulation that could be di- luted into a commercially available IV bag of normal saline (NS) for continuous intravenous infusion over 24 h. Since the human dose had not yet been established, a secondary requirement was that the formulation approach be scalable to other drug concentrations without addition of new ingredients and preferably without changes to ratios of the active and inactive ingredients.
Fig. 2. (A) Structure of pinometostat and (B) pH solubility profile of pinometostat concentration as function of pH in 200 mM citrate buffers.
The experimentally determined pH-solubility profile for pinometostat is shown in Fig. 2B. The data fit to a theoretical slope of – 1 and are gen- erally consistent with a pKa of 8.8 for the first protonation and a free base solubility in the single digit μg/mL range. Most importantly from the for- mulation development perspective, the data indicated that a pH of well under 4.0 would be needed to obtain 100 mg/mL without a solubilizer.
To avoid potential pH-related injection site issues, the decision was made to pursue a higher pH formulation. HPBCD was found to provide better and safer solubilization than other excipients evaluated.
The results of solubility studies in which both pH and HPBCD con- centration were studied are shown in Fig. 3. It can be seen that both downward pH adjustment and inclusion of HPBCD contribute to solubi- lizing the drug. Based on these studies and allowing plenty of excess sol- ubilizing capacity, a formulation containing 40% (w/v) HPBCD at pH 6.0 was chosen.
Citrate was chosen as a buffer system for two reasons. First, it is the most commonly used and best accepted buffer for that pH range in in- jectable formulations. Second, since it supplies three moles of acid per mole, citric acid could be used both to adjust the pH and to serve as a buffer in the final product without introduction of excess buffer capaci- ty. It was found that adding 80 mM of citric acid led to a product with a pH of 6.0, and this buffer strength was deemed acceptable for a product that was to be diluted and infused slowly.
Following the conclusion of preclinical studies, it was decided to initiate clinical studies with a 10 mg/mL rather than 100 mg/mL formulation. To decrease drug strength by an order of magnitude, the concentrations of all three ingredients were reduced to 1/
10th their initial levels. Thus while the 100 mg/mL product con- tains 40% (w/v) HPBCD and 80 mM citric acid, the 10 mg/mL prod- uct contains 4% (w/v) HPBCD and 8 mM citric acid. A fi nal pH of 6.0 is still obtained reproducibly without adjustment as the ratio of drug substance to HPBCD to citric acid remained constant.
The manufacturing process for the pinometostat drug product is sim- ple and scalable. In the first step, water for injection is added to a suitably sized batch kettle, and stirring is initiated. Next, citric acid and HPBCD are added to the batch kettle and allowed to dissolve. Finally, the drug sub- stance is added and allowed to dissolve before the batch is brought to vol- ume with additional water for injection. Early batches were sterilized by filtration through a 0.2 μm filter and were then aseptically filled into vials. A steam autoclave cycle has now been developed and will be used on commercial batches to provide extra sterility assurance.
The product has been tested in several simulated use studies using different compositions of IV bags and infusion sets. In all cases, the product has been found to be chemically and physically stable after preparation for dosing, and no decrease of the weight-% assay has
been observed during storage in IV bags or simulated infusion through IV administration sets.
The stability of the packaged formulation has been studied for 6 months at 40 °C/75%RH and for 24 months both at 5 °C and at 25 °C/
60%RH. Very little, if any, change has been seen in product quality attributes (appearance, pH, assay, impurities, particulates, osmo- lality). The solutions described above were used for the pharmaco- kinetic and pharmacodynamics studies discussed below.
3.2. Pharmacokinetic studies in rat and dog
The pharmacokinetics of pinometostat following single dose SC bolus administration of pinometostat drug product were evaluated in Sprague–Dawley rat and Beagle dog at doses of 1, 3 and 10 mg/kg (Fig. 4 and Table 1). In rat, absorption was rapid with plasma levels of pinometostat highest between 3 and 13 min post-dose before declining with a t1/2 of between 0.6 and 1.1 h up to 8 h post-dose. A longer termi- nal phase t1/2 was noted at the 3 and 10 mg/kg dose levels, but was not well characterized due to a lack of sampling between 8 and 24 h post- dose. This terminal phase t1/2 approximated that observed following IV administration at circa 3–5 h [21]. The mean maximum concentration (Cmax) was 242, 1046 and 1720 ng/mL following 1, 3 and 10 mg/kg doses respectively, indicating a less than dose proportional increase in Cmax between 3 and 10 mg/kg. This was associated with an increase in mean absorption time (mean residence time, oral (MRTpo) – mean residence time IV (MRTiv)) from 46 min to 103 min. The mean area under the curve to infinity (AUC0–inf) was 258, 758 and 3486 ng·h/mL, following 1, 3 and 10 mg/kg doses respectively, suggestive of a slightly higher than dose proportional increase in AUC between 3 and 10 mg/kg. The absolute SC bioavailability (F) of pinometostat in rat was complete
Fig. 4. Profile of pinometostat (EPZ-5676) mean plasma concentrations after subcutane-
Fig. 3. Solubility profile of pinometostat as a function of pH and wt-% hydroxypropyl-beta- cyclodextrin (HPBCD) solutions.
ous bolus administration in (A) SD rat and (B) beagle dog. Subcutaneous administrations: (●) 10 mg/kg pinometostat; (▲) 3 mg/kg pinometostat; (□) 1 mg/kg pinometostat.
Table 1
Summary PK parameters for pinometostat following SC administration of single doses of 1, 3 and 10 mg/kg to Sprague–Dawley rat and Beagle dog.
Species Dose (mg/kg) AUC0–last (ng·h/mL) AUC0–inf (ng·h/mL) Cmax (ng/mL) tmax (h) F (%)a t1/2 (h) MRT (h)
Rat 1 257 ± 8 258 ± 8 242 ± 26 0.05 105 ± 3 0.64 ± 0.07 1.07 ± 0.07
3 756 ± 10 758 ± 10 1046 ± 264 0.05 102 ± 1 0.60 ± 0.02 1.18 ± 0.21
10 3483 ± 596 3486 ± 596 1720 ± 350 0.217 141 ± 24 1.11 ± 0.12 2.12 ± 0.17
Dog 1 1065 ± 322 1068 ± 323 597 ± 113 0.5 119 ± 36 1.17 ± 0.01 2.26 ± 0.14
3 5250 ± 445 5264 ± 446 2790 ± 338 0.5 195 ± 17 1.21 ± 0.11 2.45 ± 0.32
10 12,761 ± 1897 12,804 ± 1906 6960 ± 1828 0.25 142 ± 21 1.21 ± 0.09 2.48 ± 0.12 All values are reported as mean ± standard deviation, with exception of tmax which is reported as the median. n = 3 per dose group.
a Bioavailability calculated based on AUC0–inf of 247 and 899 ng·h/mL obtained following a 1 mg/kg IV bolus dose in rat and dog respectively, as reported in [21].
with F ranging between 102 and 141%. Bioavailability markedly greater than 100% was determined at the 10 mg/kg dose level, reflective of the supra-proportionality in AUC observed at that dose. The nonlinearity observed could be related to saturation of an elimination mechanism or indicate the involvement of a saturable efflux process at the site of absorption.
In dog, absorption was rapid with plasma levels of pinometostat highest between 15 and 30 min post-dose before declining with a t1/2 of 1.2 h up to 8 h post-dose. Mean absorption times were short at 5, 17 and 19 min following 1, 3 and 10 mg/kg doses respectively. A longer terminal phase t1/2 was noted at all dose levels but was not well characterized due to a lack of sampling between 8 and 24 h post-dose. This terminal phase t1/2 approximated that observed following IV administration at circa 13 h [21]. The mean Cmax was 597, 2790 and 6960 ng/mL and mean AUC0–inf was 1068, 5264 and 12,804 ng·h/mL following 1, 3 and 10 mg/kg doses respectively, in- dicating reasonably dose proportional increases in exposure across the dose range. The absolute SC bioavailability of pinometostat in dog was complete and typically greater than 100%, with F ranging between 119 and 195%.
3.3. Pharmacodynamic effects of pinometostat in a subcutaneous xenograft tumor
We tested the ability of the pinometostat drug product to inhib- it a subcutaneous xenograft model of MLL-r in mice utilizing a
subcutaneous route of dosing. Pharmacokinetic properties of pinometostat dosed subcutaneously revealed exposure levels in the effi cacious range were maintained for several hours. To in- crease our exposure the maximum tolerated subcutaneous dose was determined and chosen for our effi cacy study. Treatment of MV4-11 xenograft tumors for 21 days led to signifi cant tumor growth inhibition (Fig. 5A). Both pinometostat-treated groups re- ceived 120 mg/kg/day of pinometostat. However, the group dosed t.i.d. achieved 100% mean tumor growth inhibition compared to 77% mean tumor growth inhibition achieved by the b.i.d. dosing regimen. This increased effi cacy correlated with greater body weight loss in the 40 mg/kg t.i.d group. Both the 40 mg/kg t.i.d. and 60 mg/kg b.i.d effi cacy groups had mean body weight loss, of 19 and 9% respectively. Plasma levels for both the effi cacy and pharmacokinetic/pharmacodynamic (PK/PD) cohorts were in the expected range of past pharmacokinetic studies performed in rodents.
Evaluation of H3K79 methylation was completed on MV4-11 subcutaneous xenograft tumors harvested from SCID mice follow- ing 14 days of treatment with pinometostat (Fig. 5B). Administra- tion of pinometostat through subcutaneous delivery reduced H3K79me2 in both dosed groups when compared to a vehicle treated control group. These results demonstrate that subcutane- ous dosing leads to in vivo target engagement of pinometostat as measured by the reduction of H3K79me2 in tumor tissue. Tumor tissue harvested from the pinometostat treated mice also showed
Fig. 5. Subcutaneous dosing with pinometostat leads to inhibition of tumor growth and DOT1L pharmacodynamics markers in a mouse model of MLL-rearranged leukemia. (A) Tumor growth was significantly (repeated measures ANOVA with Dunnett post-test) inhibited following 21 days of treatment with either 40 mg/kg or 60 mg/kg dosed thrice and twice daily respectively. (B) H3K79me2 levels in tumor tissue were inhibited following 14 days of subcutaneous dosing with pinometostat. H3K79me2 levels were normalized to those of total histone H3 in the same sample and are plotted as a percent of the mean H3K79me2 level in tissue from the vehicle treated (control) group, which is set at 100%. (C) Inhibition of mRNA in MLL-r target genes HOXA9 and MEIS1 as measured by qPCR in tumor tissue harvested from mice treated subcutaneously with pinometostat for 14 days. HOXA9 and MEIS1 levels are plotted as percent of vehicle control and the expression levels are normalized to the housekeeping gene B2 microglobulin.
reduction of HOXA9 and MEIS1 mRNA (Fig. 5C). Previous studies [14,15] demonstrated that inhibition of HOXA9 and MEIS1 is a di- rect consequence of DOT1L inhibition in the context of an MLL-r model.
4.Discussion
Pinometostat is a novel DOT1L inhibitor and the fi rst member of the histone methytransferase inhibitor (HMTi) class to enter clinical development as a potential therapeutic agent in MLL-r leukemia. Pinometostat is currently under clinical investigation in adult and pediatric leukemia patients bearing an MLL-rearrangement, admin- istered as CIV (studies NCT01684150 and NCT02141828 in ClinicalTrials.gov). Pinometostat showed negligible oral bioavailabil- ity in preclinical species, primarily due to its low permeability and high fi rst pass hepatic extraction. In addition, high hepatic clearance in rodents contributed to a short t1/2 [21]. Alternative modes of drug delivery for pinometostat were explored in an attempt to improve patient convenience and eliminate the technical/logistical complex- ity of CIV dosing.
Subcutaneous administration of pinometostat formulated in 0.2% (w/v) HPBCD in saline (current IV drug product) in both rat and dog re- sulted in complete bioavailability. Absorption following SC injection was also rapid in both species with Cmax observed within 30 min post- dose. In addition, mean absorption time (calculated as MRTiv subtracted from MRTpo) was short; 40 min in rat and 5 min in dog at the low dose. Rapid absorption from the SC space is commonly observed for a variety of small molecule drugs, thought to be mediated by the largely unre- stricted permeability across the vascular endothelium together with the high rate of filtration and reabsorption of fluid across the vascular capillaries (in the range of 20–40 L/day). The lymphatic system has also been implicated in the absorption and disposition of small and large molecule therapeutics administered by the SC route [22]. In addi- tion, complete SC bioavailability as compared to the previously reported low oral bioavailability also indicates that first pass hepatic extraction was largely circumvented by the SC route of administration.
The time–concentration profile following SC bolus administration obviously differs markedly from that following CIV. In our previous work, we reported the pharmacokinetics of pinometostat dosed as a CIV at 5 mg/kg/day over 7 days in rat. The exposure parameters Cmax and AUC0–t were 47 ng/mL and 4484 ng·h/mL respectively (all dose- normalized to 3 mg/kg/day to allow direct comparison with this study). The corresponding data following SC administration in rat leads to a Cmax and AUC0–t of 1046 ng/mL and 756 ng·h/mL respectively. This clearly demonstrates the difference in the exposure profile afforded by a continuous vs discrete dosing frequency, with the SC Cmax being approximately 20-fold higher and the SC AUC0–t being approximately 6-fold lower at an equivalent daily dose, and similar average concentra- tion (Cavg). The exposure and pharmacokinetics of pinometostat admin- istered by the SC route showed promise and warranted further evaluation in efficacy models of MLL-r leukemia.
The MV4-11 xenograft model of MLL-r leukemia was chosen to benchmark the activity observed with subcutaneous dosing against that obtained with continuous IV infusion. Subcutaneous treatment of MV4-11 tumor bearing mice with pinometostat led to a dose dependent reduction of tumor growth. As expected, efficacy was greater when the same daily dose was divided among 3 rather than 2 SC bolus injections. However, body weight loss was observed in both SC groups, with the t.i.d. group losing about twice as much weight as the b.i.d. group. Fol- lowing the cessation of dosing both pinometostat treated groups recov- ered body weight. To verify the mechanism of tumor inhibition, we demonstrated that both H3K79me2 and MLL-r target genes HOXA9 and MEIS1 were inhibited in excised tumor tissue, as had previously been shown with CIV dosing. Nonetheless, the activity observed through SC dosing was not enhanced over CIV delivery of pinometostat. Although the reduction of H3K79me2 and MLL-r target genes were
similar between the two routes of administration, the CIV lead to a more robust efficacy signal and had no tolerability concerns.
Based on the need for t.i.d. dosing in the xenograft studies as well as the short t1/2 and MRT observed following SC administration of the so- lution formulation, additional formulation optimization was performed in an attempt to extend the release of pinometostat from the SC space and prolong the t1/2 through extended release (i.e. flip-flop) kinetics. Thus, our primary approach to developing a fast release suspension was by particle size reduction and preparing amorphous dispersion sus- pensions. To prepare suspensions, the drug substance was first micron- ized by jet-milling. It was then suspended in medium chain triglyceride oil or an aqueous base containing suspending agents. To attain smaller particle sizes the suspensions were passed through a microfl uidizer for multiple passes at up to 18,000 lb per square inch. Median diameters in the 3–4 μm range were obtained by jet milling, but it was not possible to decrease the median diameter to less than 1 μm by micro fluidization. Particle size distributions for suspensions prepared with unmilled API, jet-milled API, and jet-milled and then micro fluidized API are shown in Fig. 6. Jet-milled API was evaluated suspended both in an aqueous ve- hicle (1% sodium carboxymethylcellulose with 0.1% polysorbate 80) and in soybean oil. SC administration of pinometostat as these suspension formulations with characterized reduced particle size did successfully extend the t1/2 of pinometostat in rat into the 10–30 h range; however, the overall systemic exposure was low and did not support further pro- gression (data not shown).
5.Conclusions
Pinometostat is a small molecule inhibitor of the DOT1L en- zyme, a histone methyltransferase that methylates lysine 79 of his- tone H3. Pinometostat is currently in Phase 1 clinical trials in relapsed refractory acute leukemia patients administered as a CIV infusion. The studies described herein investigated alternatives to CIV administration of pinometostat in order to improve patient convenience. Various sustained release technologies were consid- ered and discarded based on the required dose size as well as practical considerations. SC bolus administration of a solution for- mulation of pinometostat showed rapid absorption and complete bioavailability. SC t.i.d. dosing in xenograft models demonstrated inhibition of MLL-r tumor growth and inhibition of pharmacody- namic markers of DOT1L activity. Development of an extended re- lease formulation may prove useful in the further optimization of the SC delivery of pinometostat, moving towards a more conve- nient dosing paradigm for patients.
Fig. 6. Particle size distributions for SC suspensions of pinometostat.
Conflict of interest
All authors are current or former employees of Epizyme, Inc. with the exception of Bruce N. Rehlaender, who is a paid consultant at Pharmadirections, 5001 Weston Parkway, Suite 103, Cary, NC 27513.
Acknowledgments
We would like to thank the clinical trial principal investigators and their institutions, the employees of Epizyme, Celgene, external collabo- rators and most importantly, the patients participating in the clinical trials.
References
[1]J.C. Byrd, K. Mrozek, R.K. Dodge, A.J. Carroll, C.G. Edwards, D.C. Arthur, M.J. Pettenati, S.R. Patil, K.W. Rao, M.S. Watson, P.R. Koduru, J.O. Moore, R.M. Stone, R.J. Mayer, E.J. Feldman, F.R. Davey, C.A. Schiffer, R.A. Larson, C.D. Bloomfield, Pretreatment cytoge- netic abnormalities are predictive of induction success, cumulative incidence of re- lapse, and overall survival in adult patients with de novo acute myeloid leukemia: results from Cancer and Leukemia Group B (CALGB 8461), Blood 100 (2002) 4325–4336.
[2]J.L. Hess, MLL: a histone methyltransferase disrupted in leukemia, Trends Mol. Med. 10 (2004) 500–507.
[3]A.V. Krivtsov, S.A. Armstrong, MLL translocations, histone modifications and leukae- mia stem-cell development, Nat. Rev. Cancer 7 (2007) 823–833.
[4]R.K. Slany, The molecular biology of mixed lineage leukemia, Haematologica 94 (2009) 984–993.
[5]T. Nakamura, T. Mori, S. Tada, W. Krajewski, T. Rozovskaia, R. Wassell, G. Dubois, A. Mazo, C.M. Croce, E. Canaani, ALL-1 is a histone methyltransferase that assembles a supercomplex of proteins involved in transcriptional regulation, Mol. Cell 10 (2002) 1119–1128.
[6]M.J. Chang, H. Wu, N.J. Achille, M.R. Reisenauer, C.W. Chou, N.J. Zeleznik-Le, C.S. Hemenway, W. Zhang, Histone H3 lysine 79 methyltransferase Dot1 is required for immortalization by MLL oncogenes, Cancer Res. 70 (2010) 10234–10242.
[7]T.A. Milne, S.D. Briggs, H.W. Brock, M.E. Martin, D. Gibbs, C.D. Allis, J.L. Hess, MLL tar- gets SET domain methyltransferase activity to Hox gene promoters, Mol. Cell 10 (2002) 1107–1117.
[8]Y. Okada, Q. Feng, Y. Lin, Q. Jiang, Y. Li, V.M. Coffield, L. Su, G. Xu, Y. Zhang, hDOT1L links histone methylation to leukemogenesis, Cell 121 (2005) 167–178.
[9]S.C. Monroe, S.Y. Jo, D.S. Sanders, V. Basrur, K.S. Elenitoba-Johnson, R.K. Slany, J.L. Hess, MLL-AF9 and MLL-ENL alter the dynamic association of transcriptional regula- tors with genes critical for leukemia, Exp. Hematol. 39 (2011) 77–86 (e71-75).
[10]D. Mueller, C. Bach, D. Zeisig, M.P. Garcia-Cuellar, S. Monroe, A. Sreekumar, R. Zhou, A. Nesvizhskii, A. Chinnaiyan, J.L. Hess, R.K. Slany, A role for the MLL fusion partner
ENL in transcriptional elongation and chromatin modification, Blood 110 (2007) 4445–4454.
[11]Q. Feng, H. Wang, H.H. Ng, H. Erdjument-Bromage, P. Tempst, K. Struhl, Y. Zhang, Methylation of H3-lysine 79 is mediated by a new family of HMTases without a SET domain, Curr. Biol. 12 (2002) 1052–1058.
[12]D.J. Steger, M.I. Lefterova, L. Ying, A.J. Stonestrom, M. Schupp, D. Zhuo, A.L. Vakoc, J.E. Kim, J. Chen, M.A. Lazar, G.A. Blobel, C.R. Vakoc, DOT1L/KMT4 recruitment and H3K79 methylation are ubiquitously coupled with gene transcription in mammalian cells, Mol. Cell. Biol. 28 (2008) 2825–2839.
[13]S.R. Daigle, E.J. Olhava, C.A. Therkelsen, A. Basavapathruni, L. Jin, P.A. Boriack-Sjodin, C.J. Allain, C.R. Klaus, A. Raimondi, M.P. Scott, N.J. Waters, R. Chesworth, M.P. Moyer, R.A. Copeland, V.M. Richon, R.M. Pollock, Potent inhibition of DOT1L as treatment of MLL-fusion leukemia, Blood 122 (2013) 1017–1025.
[14]S.R. Daigle, E.J. Olhava, C.A. Therkelsen, C.R. Majer, C.J. Sneeringer, J. Song, L.D. Johnston, M.P. Scott, J.J. Smith, Y. Xiao, L. Jin, K.W. Kuntz, R. Chesworth, M.P. Moyer, K.M. Bernt, J.C. Tseng, A.L. Kung, S.A. Armstrong, R.A. Copeland, V.M. Richon, R.M. Pollock, Selective killing of mixed lineage leukemia cells by a potent small- molecule DOT1L inhibitor, Cancer Cell 20 (2011) 53–65.
[15]K.M. Bernt, N. Zhu, A.U. Sinha, S. Vempati, J. Faber, A.V. Krivtsov, Z. Feng, N. Punt, A. Daigle, L. Bullinger, R.M. Pollock, V.M. Richon, A.L. Kung, S.A. Armstrong, MLL- rearranged leukemia is dependent on aberrant H3K79 methylation by DOT1L, Can- cer Cell 20 (2011) 66–78.
[16]A.T. Nguyen, O. Taranova, J. He, Y. Zhang, DOT1L, the H3K79 methyltransferase, is re- quired for MLL-AF9-mediated leukemogenesis, Blood 117 (2011) 6912–6922.
[17]A. Basavapathruni, L. Jin, S.R. Daigle, C.R. Majer, C.A. Therkelsen, T.J. Wigle, K.W. Kuntz, R. Chesworth, R.M. Pollock, M.P. Scott, M.P. Moyer, V.M. Richon, R.A. Copeland, E.J. Olhava, Conformational adaptation drives potent, selective and dura- ble inhibition of the human protein methyltransferase DOT1L, Chem. Biol. Drug Des. 80 (2012) 971–980.
[18]W. Yu, E.J. Chory, A.K. Wernimont, W. Tempel, A. Scopton, A. Federation, J.J. Marineau, J. Qi, D. Barsyte-Lovejoy, J. Yi, et al., Catalytic site remodelling of the DOT1L methyltransferase by selective inhibitors, Nat. Commun. 3 (2012) 1288.
[19]L. Chen, A.J. Deshpande, D. Banka, K.M. Bernt, S. Dias, C. Buske, E.J. Olhava, S.R. Daigle, V.M. Richon, R.M. Pollock, S.A. Armstrong, Abrogation of MLL-AF10 and CALM-AF10- mediated transformation through genetic inactivation or pharmacological inhibition of the H3K79 methyltransferase Dot1l, Leukemia 27 (2013) 813–822.
[20]A.J. Deshpande, L. Chen, M. Fazio, A.U. Sinha, K.M. Bernt, D. Banka, S. Dias, J. Chang, E.J. Olhava, S.R. Daigle, V.M. Richon, R.M. Pollock, S.A. Armstrong, Leukemic transfor- mation by the MLL-AF6 fusion oncogene requires the H3K79 methyltransferase Dot1l, Blood 121 (2013) 2533–2541.
[21]A. Basavapathruni, E.J. Olhava, S.R. Daigle, C.A. Therkelsen, L. Jin, P.A. Boriack-Sjodin, C.J. Allain, C.R. Klaus, A. Raimondi, M.P. Scott, A. Dovletoglou, V.M. Richon, R.M. Pollock, R.A. Copeland, M.P. Moyer, R. Chesworth, P.G. Pearson, N.J. Waters, Nonclin- ical pharmacokinetics and metabolism of EPZ-5676, a novel DOT1L histone methyl- transferase inhibitor, Biopharm. Drug Dispos. 35 (2014) 237–252.
[22]D.N. McLennan, C.J. Porter, S.A. Charman, Subcutaneous drug delivery and the role of the lymphatics, Drug Discov. Today Technol. 2 (2005) 89–96.