Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Tumor necrosis factor- (TNF-) is a pleiotropic cytokine that signals through multiple kinase pathways to regulate gene expression and a wide range of cellular processes. TNF- overproduction is implicated in many inflammatory diseases (Beutler, 1999). Rheumatoid arthritis treatment has been advanced by TNF- inhibitors (Maini et al., 1998, 1999; Moreland et al., 1999; Weinblatt et al., 1999; Bathon et al., 2000; Lipsky et al., 2000; Schattenkirchner et al., 2000; Van de Putte et al., 2000), as TNF- plays a pivotal role in both rheumatoid inflammation (Choy and Panayi, 2001) and the erosion of adjacent bone (Tak and Bresnihan, 2000). TNF- inhibitors are additionally indicated for the induction and maintenance of remission, and fistula closure in Crohn's disease (Targan et al., 1997; Present et al., 1999).

ISIS 104838 is a 20-base phosphorothioate (PS) oligonucleotide identified through cell-based screening. In addition to the PS modification, ISIS 104838 contains 2'-O-(2-methoxyethyl) modified (2'-MOE) nucleosides at each of the five terminal 3' and 5' nucleotide sugars, and is considered a second generation "chimeric" chemistry. The central ten PS oligonucleotides are unmodified 2' deoxyribose nucleotides and are termed the "gap" (Dean et al., 2001). Antisense oligonucleotides hybridize to their target messenger RNA (mRNA) through Watson-Crick base pair interactions, offering a very high level of target specificity. The mRNA in a RNA:DNA duplex is cleaved by the ubiquitous nuclease RNase H (Crooke, 1999). This function is effectively supported by the ten deoxyribonucleotides within the central gap, but not by the 2'-MOE modified nucleotides (Baker and Monia, 1999). The 2'-MOE modified nucleotides increase binding affinity for the target mRNA and resist endogenous exonucleases, thereby prolonging drug half-life (Bennett et al., 2000; Henry et al., 2000; Yu et al., 2001).

We performed two phase I trials evaluating the safety and pharmacokinetics of ISIS 104838 by i.v. or s.c. administration, representing the first report of human administration of a 2'-MOE modified antisense oligonucleotide. An ex vivo analysis of efficacy was also performed following i.v. dosing.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Identification and Characterization of ISIS 104838

Oligonucleotide screening was performed using neonatal human epidermal keratinocytes (Cascade Biologics, Portland, OR) seeded at nonconfluence in 96-well plates (Becton Dickinson, Franklin Lakes, NJ). Cells were treated with oligonucleotides at 200 nM in 6 µg/ml Lipofectin (Invitrogen, Carlsbad, CA) for 4 h at 37°C. Medium was replaced and cells were incubated for 2 h in the presence or absence of 100 nM phorbol myristate acetate (PMA) (Sigma-Aldrich, St. Louis, MO). Poly A+ RNA was isolated from cell lysates using a 5'-biotin-labeled 2'-MOE modified oligonucleotide (U)₂₀ capture probe in combination with a NeutrAvidin capture plate (Pierce, Rockford IL). TNF- mRNA was measured by real-time quantitative RT-PCR on the ABI Prism 7700 system (Applied Biosystems Inc., Foster City, CA) as recommended by the manufacturer (Gibson et al., 1996; Winer et al., 1999).

A dose response analysis was performed in keratinocytes treated with oligonucleotides at 30, 100, or 300 nM with Lipofectin for 4 h at 37°C. Medium was replaced and cells incubated in the presence or absence of 100 nM PMA for 12 h; then medium was collected for quantitation of TNF- protein by ELISA (TNF- DuoSet kit, R&D Systems, Minneapolis, MN).

ISIS 104838 activity and specificity were evaluated in human THP-1 monocytic cells (American Type Culture Collection, Manassas VA). Cells were transfected with oligonucleotide by eletroporation (BTX Electro Cell Manipulator 600, San Diego CA) and then recovered in RPMI-1640 medium with 10% fetal bovine serum (Invitrogen) for 5 h at 37°C. Lipopolysaccharide (LPS) at 100 ng/ml was added for 2 h. Total RNA was isolated using a RNeasy MiniKit (QIAGEN, Valencia, CA) and specific mRNA levels were determined by real-time RT-PCR.

Primers and probes used for real-time RT-PCR analysis were designed using the Primer Express software (Applied Biosystem Inc.). Relative amounts of target mRNA per well were determined by the standard curve method (Winer et al., 1999). G3PDH mRNA levels were used as an internal standard.

Drug Product

ISIS 104838 is a racemic mixture. The percentage of full length ISIS 104838 in the drug product was 93.2%, with the major impurity consisting of n-1 deletion sequences (1.8%), and 94% of the full-length oligonucleotide was fully thioated. Dosing was performed based on the quantity of full-length 20-mer. ISIS 104838 for i.v. infusion was provided by Isis Pharmaceuticals, Inc., as an isotonic, sterile 10 mg/ml solution in phosphate-buffered saline (pH 7.6). The injectates were within specified limits for bacterial endotoxin (<5 endotoxin units/ml for the 10 mg/ml, and <100 endotoxin units/ml for the 200 mg/ml solutions). The i.v. drug or placebo (0.9% saline) was administered by 1 h infusion in 100 ml of sterile 0.9% saline. The research pharmacist was not blinded.

ISIS 104838 for s.c. injection was provided by Isis Pharmaceuticals, Inc. as a 200 mg/ml solution in sterile water, adjusted with hydrochloric acid and sodium hydroxide to pH 7.4. The drug product lot was identical to that utilized for infusions. For single dose subjects, the drug was administered in a single 1-ml injection using rising ISIS 104838 concentrations of 25 to 200 mg/ml, diluted as necessary with sterile 0.9% saline by the research pharmacist, who was not blinded. The multiple dose subjects received varying volumes of the ISIS 104838 200 mg/ml solution to achieve a total dose of 0.1 to 6 mg/kg. The maximum injection volume was 1 ml; therefore, some subjects received two or three injections to achieve their calculated dose. All injections were administered s.c. using a 27-gauge needle into the abdominal wall in a rotational schema. The placebo was sterile 0.9% saline with riboflavin to provide appropriate color in matching vials.

Subjects

Healthy males aged 18 to 45 years were dosed at Guy's Drug Research Unit (GDRU), with local ethics committee approval, and the studies were carried out in accordance with the Declaration of Helsinki. Subjects underwent full medical history, physical examination, and laboratory testing including drug screens, hepatitis A, B, C, and human immunodeficiency virus I/II testing. Subjects were admitted the night before treatment after abstaining from alcohol for 72 h and remained on the unit for 24 h following every dose. They received a light breakfast at least 1 h prior to dosing and refrained from smoking, caffeine-containing drinks, and grapefruit juice for 1 h before and 12 h following each dose. They remained recumbent, with continuous EKG recording, for 4 h following the completion of each infusion or injection.

ISIS 104838 Schedule

Intravenous Dosing. Four volunteers were recruited into each i.v. cohort, receiving doses of 0.1, 0.5, 1, 2, 4, or 6 mg/kg on days 1, 8, 10, and 12. One subject in each cohort was randomly assigned to receive placebo. Dose escalation was made at intervals of at least 1 week, after an assessment of safety for the previous cohort. An additional obese cohort of two subjects subsequently received 2 mg/kg dosing. These subjects weighed more than 10% above ideal body weight (ideal body weight = 48 + 1.1 kg for every centimeter in height over 152 cm) and had 25% body fat by the Harpenden skinfold caliper method. Body mass index was calculated by weight (kg)/height (m)²; by the National Institutes of Health guidelines obesity is defined as body mass index 30.

Subcutaneous Dosing. In the single dose trial, four volunteers were recruited into each of four cohorts, and received a single 1-ml injection at a concentration of 25, 50, 100, or 200 mg/ml. One subject in each cohort was randomly assigned to placebo. Dose escalation was made at intervals of at least 1 week, after assessment of safety for the previous cohort. The single doses were evaluated prior to selecting the 200 mg/ml ISIS 104838 concentration for the multidose cohorts. In these cohorts, four volunteers were recruited into each of six cohorts, and received 0.1, 0.5, 1, 2, 4, or 6 mg/kg s.c. on days 1, 3, and 5. One subject in each cohort was randomly assigned to placebo.

Safety Monitoring

In both studies, safety labs included hematology, blood chemistry, urinalysis, coagulation (aPTT, PT, and international normalized rate) and complement split product (C3a). Blood pressure and heart rate were monitored before dosing and at intervals after dosing. All laboratory tests, including virology and drug screens, were performed at GDRU. Local side effects at injection sites were assessed at 2, 12, and 24 h, and on days 4 and 7 after the single s.c. injections. The multidose subjects had an evaluation of injection sites at 2, 12, and 24 h with each injection, and on days 11 and 14.

The two i.v. subjects who developed rashes had total IgE levels and viral titers (IgM and IgG) assayed for cytomegalovirus, mumps, measles, herpes simplex virus type 1, herpes simplex virus type 2, and Epstein-Barr virus (including Epstein-Barr nuclear antigen) using plasma obtained following treatment cessation. These tests, and an erythrocyte sedimentation rate for one of the subjects, were performed at GDRU. Plasma cytokine levels using frozen EDTA plasma samples for these subjects were also assayed on day 1 prior to treatment and again following the completed infusion on day 8 (Huntingdon Life Sciences, Huntingdon, UK) by Quantitine ELISA kits (R&D Systems Europe, Abingdon, UK). The lower limit of detection for TNF- was 31.2 pg/ml (kit DTA50), 7.8 pg/ml for IL-1 (kit DLB50), and 6.25 pg/ml for IL-6 (kit D6050).

Ex Vivo Analysis of Cytokine Production

Whole blood drawn in a lithium heparin tube was cultured for each i.v. subject at GDRU in a 24-well plate. The assay was performed with a predosing sample and a second sample from 24 h after the final day 12 infusion. Samples (8 duplicates; 10 ml each) were incubated at 38°C with agitation for 4 h in the presence of LPS (Sigma-Aldrich), at a final concentration of 0, 0.1, 1.0, and 10 ng/ml in Hanks' balanced salt solution (Invitrogen). Cell-free supernatants were pooled, frozen at 80°C and shipped to Huntingdon Life Sciences, where they were assayed using the ELISA kits. Validation assays confirmed no assay interference by ISIS 104838. Standards were assayed in duplicate and samples were assayed once at each dilution. The day 13 data were expressed as a percentage of the day 1 control results.

ISIS 104838 Pharmacokinetics

Plasma concentrations were measured at Isis Pharmaceuticals, Inc. by a hybridization-dependent nuclease ELISA method (lower limit of quantitation = 0.77 ng/ml, with a 25% coefficient of variation). Plasma was aliquoted into wells containing a complementary 20-oligomer hybridization probe. The quality control (QC) standards run with each ELISA were duplicates of three known ISIS 104838 concentrations (low, medium, and high), in drug-free plasma samples (accuracy 100 ± 20% and precision 20% for all concentrations above the lower limit of quantitation). Stability of storage and handling conditions were fully evaluated for up to 6 months. The assay was validated in accordance with FDA guidelines. Negative controls (ISIS 104838 mismatches and metabolites) were used for validation of assay selectivity and specificity but were not run concurrently. Cross reactivity was less than 1% for mismatches and metabolites.

Day 1 ISIS 104838 plasma levels were also measured at Isis Pharmaceuticals, Inc. by capillary gel electrophoresis (CGE) as previously described (lower limit of quantitation = 0.154 µg/ml, with a 25% coefficient of variation) (Leeds et al., 1996). QC standards of ISIS 104838 in drug-free reference plasma samples were run with each assay.

Urinary concentrations were measured at Isis Pharmaceuticals, Inc. by a CGE assay (lower limit of detection = 0.039 µg/ml, with a 25% coefficient of variation) as described above. Calibration standards and three QC standards were prepared in drug-free urine containing known amounts of ISIS 104838.

ISIS 104838 metabolites in urine were analyzed by CGE. CGE peak areas were normalized using the following equation: nA₂ = (A₂/T_m2)  (A₁/T_m1), where nA₂ = normalized peak area of analyte; A₂ = peak area of analyte; T_m2 = CGE migration time of analyte; A₁ = peak area of internal standard; and T_m1 = CGE migration time of internal standard. Normalized peak areas of calibration standards were regressed linearly (using 1/x weighting) to define the peak response-concentration relationship. Concentrations of ISIS 104838 in plasma and urine were determined from matrix-specific calibration curves. Concentrations of major metabolites (n-8 to n-12) were also calculated and normalized using the extinction coefficients for standard oligonucleotides of equivalent length.

Area under the plasma concentration-time curve (AUC) was calculated using the linear trapezoidal rule. For s.c. dosing, the apparent distribution half-life (t_1/2) was calculated using 0.693/K_distribution (K = elimination rate constant). For i.v. dosing, volume of distribution at steady-state (V_SS) and mean residence time (MRT) were calculated using statistical moment theory, and the effective plasma disposition half-life (t_1/2eff) was estimated using MRT multiplied by 0.693 (Perrier, 1982). The first-order rate of disposition from plasma (_Z) during the apparent terminal phase was estimated using nonlinear regression analysis (WinNonlin, version 3.1) (Pharsight, Mountain View, CA) on the last three or more samples following drug administration. The apparent terminal disposition half-life from plasma was estimated using (0.693/_Z). AUC was extrapolated to infinity by dividing the final measurable concentration (C_last) by the slope of the disposition curve (_Z). Plasma clearance (CL) was calculated (dose/AUC).

The amount of ISIS 104838 excreted during the first 24 h after dosing was calculated by multiplying the total oligonucleotide or intact 20-mer ISIS 104838 concentrations in each urinary aliquot with the volume and summing all collection period results. The fraction of the administered dose that was excreted was calculated by dividing the total amount of oligonucleotide in urine by the total administered dose.

Compartmental modeling was performed for all i.v. subjects with plasma monitoring through 94 days after the final dose (WinNonlin, version 3.1). One-, two-, and three-compartment models were evaluated. The apparent terminal elimination half-life was determined by dividing 0.693 (ln(2)) by the apparent terminal elimination rate constant (_Z).

Statistical Analysis

Placebo and treated subjects were compared for laboratory and safety parameters using a two-tailed Student's t test. Differences in pharmacokinetic parameters for the individual dose cohorts were compared using an unpaired two-tailed Student's t test.

LPS-stimulated ex vivo cytokine production was compared for each patient before and after ISIS 104838 treatment. The trend for the ISIS 104838 dose effect was evaluated by calculating a slope for each subject's cytokine production on day 1, stimulated by placebo or one of three LPS concentrations. This slope was compared with the slope for the subject's cytokine production curve on day 13. The difference of the two slopes, and the difference in AUC below each of the two slopes, was correlated to ISIS 104838 dose level using a simple regression.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Identification of ISIS 104838

ISIS 104838 was identified in a screen of 264 phosphorothioate 2'-MOE modified chimeras, 20 nucleotides in length, which spanned TNF- mRNA. A 10 nucleotide 2'-deoxy gap was used to maximize RNase H activity, and to concomitantly maximize the target mRNA affinity as well as the nuclease resistance provided by the 2'-MOE modifications (McKay et al., 1999; Wu et al., 1999; Zhang et al., 2000). The ISIS 104838 sequence is GCTGATTAGAGAGAGGTCCC (2'-MOEs underlined). This sequence shows no full homology for other mRNA in the NCBI human genome sequence database. All cytosines were 5'-methylated to minimize lymphocyte mitogenicity (Boggs et al., 1997). Dose response analysis showed ISIS 104838 to be the most potent inhibitor of TNF- in PMA-activated keratinocytes, reducing secreted TNF- protein levels at 300 nM to 15% of the levels produced in the absence of oligonucleotide (Fig. 1). The location of the 10 nucleotide 2'-deoxy gap was optimized by repositioning the gap, in one base increments, while maintaining a constant nucleotide sequence. The 5-10-5 MOE structure of ISIS 104838 provided the most complete suppression of TNF- protein production. ISIS 104838 showed an IC₅₀ value <1 µM for inhibition of TNF- mRNA in LPS-activated THP-1 monocytic cells (Fig. 2). This response was sequence-specific. A control oligonucleotide with five mismatches for the human TNF- gene did not reduce TNF- mRNA levels. In addition, ISIS 104838 had a minimal effect on IL-1 mRNA production, confirming that ISIS 104838 was selective for TNF- (Fig. 2).

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Dose response for ISIS 104838 and other 2'-MOE modified oligonucleotide lead candidates in PMA-activated human keratinocytes. Secreted TNF- protein levels after oligonucleotide treatment at 30, 100, or 300 nM in triplicate is displayed as a percentage of the untreated baseline. ISIS 104838 at 300 nM shows maximal TNF- protein suppression, to 15% of baseline. 5MM control is ISIS 16798, a 5-10-5 MOE gapmer that complements mouse TNF- but has five mismatches for human TNF-.

View larger version (34K): [in this window] [in a new window]   Fig. 2.   Specificity of ISIS 104838 in LPS-induced THP-1 monocytic cells. RT-PCR analysis of ISIS 104838-treated cells, relative to mRNA isolated from untreated cells, demonstrates suppression of TNF- mRNA levels after LPS stimulation to <10% but no suppression of IL-1 mRNA levels. The control oligonucleotide is ISIS 16798, a 5-10-5 MOE gapmer that complements mouse TNF- but has five mismatches for human TNF-.

Subjects

Twenty-six males (23 Caucasian, 1 Hispanic, 1 Black, 1 mixed race) aged 18 to 44 (mean 26.7 ± 6.4 years) enrolled in the i.v. trial. The 24 nonobese subjects weighed 59.8 to 86.9 kg (mean 74.9 ± 7.0 kg) and had an ideal body weight (IBW) of 61.2 to 89.8 kg (mean 75.7 ± 7.1 kg); three subjects were minimally above IBW + 10%. The two obese subjects weighed 121.1 and 134.9 kg, with IBW of 73.3 and 86.5 kg, and body mass indexes of 39.5 and 38.6. Sixteen (15 Caucasian, 1 mixed race) males aged 19 to 33 enrolled in the s.c. single dose trial. Twenty-four (23 Caucasian, 1 mixed race) males aged 21 to 35 enrolled in the s.c. multidose trial. The s.c. patients had a mean age of 25.4 (± 4.3) years, weighed 55.5 to 88.7 kg (mean 72.8 ± 7.1 kg), and had an IBW range of 59 to 86.5 kg (mean 74.5 ± 6.0 kg); 6 subjects were minimally outside IBW ± 10%. Three subjects were withdrawn by the investigator after two doses for safety considerations (one each at 4 mg/kg i.v. and 6 mg/kg i.v., and one at 6 mg/kg s.c.).

Safety

Transient and reversible aPTT prolongation was observed at the completion of the higher dose i.v. infusions and, to a much lesser degree, following the highest dose s.c. injections. The maximum aPTT change occurred at the end of each infusion (EOI) and averaged 5, 7, 11, 27, and 33 s, respectively, for 0.5 mg/kg to 6 mg/kg dose cohorts. Transient aPTT prolongation after i.v. infusions correlated linearly with the maximal plasma concentration (C_max) at EOI, occurring at a rate of 0.63 s of aPTT prolongation per 1 µg/ml increment in C_max (r = 0.94 by Pearson correlation). This resolved within 2 to 4 h after EOI at 2 mg/kg and by 5 h after EOI for the 4 and 6 mg/kg subjects (Fig. 3). There was no evidence of compromised hemostasis. There were no detectable aPTT changes following s.c. injections below 2 mg/kg. The maximum aPTT change occurred at 3 h following a 2, 4, or 6 mg/kg injection and averaged 3, 3.7, and 7.8 s, respectively. There were corresponding PT changes for i.v. infusions at 2 mg/kg, peaking with an average PT prolongation of 3.8 s following a 6 mg/kg infusion. There were no dose-related changes in C3a.

View larger version (25K): [in this window] [in a new window]   Fig. 3.   Time course of transient aPTT prolongation following ISIS 104838 infusion in non-obese normal volunteers. aPTT normalization occurred within 2 to 4 h at doses 2 mg/kg, and by 5 to 6 h after 4 or 6 mg/kg.

No serious or severe adverse events were reported. The two i.v. subjects that received only two doses developed dyshidrotic eczema with moderate knee arthralgia (4 mg/kg dose) in one case, and palmar erythema with mild skin peeling (6 mg/kg dose) in the other case. The subject with arthralgia had an erythrocyte sedimentation rate of 3 at 9 days following the final ISIS 104838 dose. These two subjects had normal baseline and day 8 post-treatment plasma levels of IL-6, IL-1, and TNF-. Total IgE levels were normal and viral antibody titers were negative or IgG (no IgM) when tested on day 10 for one and on day 21 for the other. The one s.c. subject who received only two doses experienced a reversible platelet decrease from 191 to 113 × 10⁹/liter. No particular adverse event occurred with disproportionate frequency after active treatment, and all reported adverse events were mild with the exception of the arthralgia. Headache was the most commonly reported adverse event and was considered possibly related to ISIS 104838 in one i.v. subject and in two cases with onset at 6 to 7 h following a s.c. injection.

Skin injection sites showed some mild tenderness, erythema or induration, typically developing at 12 to 24 h and resolving by 4 days post dosing. This was more prominent with the 200 mg/ml concentration and 6 mg/kg injections. No subject developed local adenopathy. One subject incorrectly received a 4 mg/kg s.c. dose in a volume of 1.8 ml at a single abdominal location, without complication.

No changes in EKGs or vital signs were observed. The median percent change from baseline in laboratory safety variables revealed no dose-related differences between drug and placebo-treated subjects, with the exception of the described aPTT changes.

Pharmacokinetics

Intravenous Dosing. Peak plasma concentrations (C_max) following i.v. infusions were seen at the EOI and were dose-proportional over the studied dose range (Table 1). There was minimal variability between subjects at a single dose. Plasma concentrations decreased rapidly after EOI (Fig. 4). By 4 to 6 h after EOI, plasma concentrations were 10- to 20-fold lower. There were no shortened oligonucleotide metabolites observed in plasma by the 2-h time point, and minimal evidence of metabolites in the 0- to 2-h urinary aliquot, suggesting that ISIS 104838 is metabolically stable prior to cellular uptake.

View larger version (28K): [in this window] [in a new window]   Fig. 4.   Pharmacokinetic results following single intravenous ISIS 104838 doses. Peak concentrations were dose-proportional, and plasma levels fell rapidly following the end of an infusion (T = 0). The lower limit of quantitation (LLOQ) for plasma samples is highlighted for two methodologies (CGE, ELISA, see Materials and Methods for details); three subjects per time point.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Presence of metabolites by CGE in plasma and urine following s.c. and i.v. ISIS 104838 administration. Plasma values at 3 h are compared with urine aliquots collected from 2 to 4 h following infusion or injection. I.S. = internal standard; OD260 absorbance units on the y-axis; relative RT = relative running time (ratio to internal standard's retention time).

View larger version (22K): [in this window] [in a new window]   Fig. 6.   Pharmacokinetic results in a single subject following repetitive intravenous ISIS 104838 dosing at 6 mg/kg. The final sample was obtained 80 days after the final dose. The calculated terminal elimination half-life is 26.8 days.

Subcutaneous Dosing. Following a single s.c. injection, the time to C_max (T_max) ranged between 1.5 and 4.7 h, depending upon dose and concentration (Table 2). C_max was dose-proportional over the studied dose range. Plasma concentrations decreased to 10- to 20-fold lower than C_max by 12 h after an injection.

View larger version (18K): [in this window] [in a new window]   Fig. 7.   Comparison of ISIS 104838 plasma concentrations by s.c. injection versus 1 h i.v. infusion for similar dose levels (200 mg/ml s.c. 2.5 mg/kg). Time is expressed as hours following the completion of injection or infusion. Cmax is lower and plasma levels fall more gradually following s.c. injections; three subjects per time point.

View larger version (35K): [in this window] [in a new window]   Fig. 8.   Reproducible pharmacokinetics for ISIS 104838 by subcutaneous injection. ISIS 103838 was administered on 3 protocol days (days 1, 3, and 5) and plasma concentrations at 3 h were averaged for 3 subjects per dose level. The maximum single injection volume was 1 ml (200 mg/ml), and subjects in the higher dose cohorts received two or three injections to achieve the correct milligram per kilogram dosage. No dose accumulation was noted with repetitive dosing.

Stimulated Cytokine Production ex Vivo

The level of TNF- production following LPS stimulation ex vivo was diminished for each cohort in a dose-dependent manner following treatment with ISIS 104838 (Fig. 9). Spontaneous baseline TNF- production without LPS stimulation was detected in only one subject. One subject's day 13 sample was clotted, and the assay was not performed. Labeling errors appeared to reverse the ascending LPS concentration in two 1 mg/kg subjects (lowest TNF- production after LPS 10 ng/ml; highest production after no LPS), and these two subjects were excluded from the analysis.

View larger version (21K): [in this window] [in a new window]   Fig. 9.   Stimulated TNF- production ex vivo. The effect of ISIS 104838 treatment on the ability to stimulate TNF- production by whole blood during a 4-h LPS exposure. Samples were obtained at pretreatment on day 1 and again on day 13 at 24 h following the final ISIS 104838 infusion (4). A curve was generated for TNF- production over a range of LPS concentrations (0, 0.1, 1, and 10 ng/ml) on each day, and the difference between slopes is plotted for each subject. The ability to produce TNF- is limited by the presence of ISIS 104838, in a dose-responsive manner (p = 0.0087).

Each subject's LPS-stimulated TNF- production on day 13 was expressed as a percentage of the pretreatment day 1 value. Variability for the assay was observed in the LPS-stimulated cultures, without study drug treatment. On day 13, placebo subjects ranged from 26 to 260% (mean 118%) of basal levels after 1 ng/ml LPS stimulation, and from 19 to 109% (mean 78%) of basal levels after 10 ng/ml LPS stimulation.

In the 6 mg/kg ISIS 104838 cohort, TNF- levels were an average 43 and 55% lower than levels for placebo patients, respectively, following 1 and 10 ng/ml LPS stimulation. The dose-ranging effect of increasing ISIS 104838, across all levels of LPS stimulation (0, 0.1, 1, and 10 ng/ml), showed a significant inhibition of TNF- production (p = 0.0087 by mixed effect model). A comparison of area under the curve for the day 1 and day 13 curves also showed a significant decrease in TNF- production with increasing ISIS 104838 dose (p = 0.0056).

The effect of ISIS 104838 on the ex vivo production of IL-6 and IL-1 was evaluated. LPS increased both IL-6 and IL-1 levels. There was a trend for decreasing IL-6 production (p = 0.09) but no decrease in IL-1 with increasing ISIS 104838 dose.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

This is the first report of a chimeric, 2'-O-(2-methoxyethyl)-modified phosphorothioate oligonucleotide administered to human volunteers and overall supports good tolerability by both s.c. and i.v. routes. The anticipated advantages of this chemistry include greater potency, well tolerated subcutaneous delivery, and greater stability allowing a less frequent dosing interval (Dean et al., 2001; Yu et al., 2001). The antisense approach to inhibiting gene expression at the mRNA level has recently been reviewed (Stein, 2001).

An assessment of pharmacologic effect was undertaken in these healthy subjects, with nondetectable basal TNF- levels, by evaluating their peripheral blood leukocyte production of TNF- after stimulation. The level of TNF- production following LPS coculture ex vivo was diminished for each cohort in a dose-dependent manner following i.v. treatment with ISIS 104838 (p < 0.01). Because the day to day variability for individual placebo subjects was high in the setting of LPS stimulation, this method can only provide an approximate estimation of the physiologic effects. Future studies will be required to evaluate the effect of ISIS 104838 in vivo in patients with inflammatory conditions and pathological TNF- elevations.

Significant exposure to ISIS 104838 was achieved with relatively small pharmacokinetic variability at any dose level. Maximum plasma concentrations were dose proportional and predictable. At the higher i.v. doses, oligonucleotide distribution into tissues such as liver and kidney becomes saturated, resulting in decreased plasma clearance rates (Henry et al., 2001). The plasma concentration-time profile showed a rapid plasma distribution phase over the first 4 to 6 h and subsequently at least one additional slower disposition phase. This second phase may more directly reflect the true elimination rate of ISIS 104838 as it exits the distributed tissues. Clearance and half-life values from target tissues have not been evaluated in human subjects, but would be anticipated from animal studies to show significantly slower elimination rates (days versus hours) compared to the plasma distribution rates (Bennett et al., 2000; Levin et al., 2001; Yu et al., 2001). Terminal elimination rate constants in plasma supported a terminal half-life of approximately 14 to 25 days in this study, similar to the tissue elimination half-lives determined in rats and monkeys (10-30 days, depending on tissue). The longer terminal elimination half-life observed at 6 mg/kg is likely due to the ability to detect low ISIS 104838 concentrations in plasma at later time points.

Only a small fraction of the full-length parent drug was excreted in urine following either route of administration, likely due to high plasma protein binding, which minimizes renal loss (Dean et al., 2001; Geary et al., 2001b). ISIS 104838 is ultimately excreted from the body following endonuclease digestion and urinary excretion in mouse, rat, and monkey (Cummins et al., 2001; Geary et al., 2001a), because the shorter oligonucleotide metabolites have lower plasma protein binding affinity. The data from these studies suggest a similar metabolism profile for humans. Nevertheless, the residence time in tissue is long, as evidenced by the long terminal elimination half-life. These pharmacokinetic data support alternate week or monthly dosing, although the residence half-life in specific tissues will vary, as the 14- to 25-day elimination half-life represents an integration of many organ elimination curves (Yu et al., 2001).

Obese subjects attained a higher C_max and AUC following 2 mg/kg i.v. dosing, compared with non-obese subjects, most likely due to the limited distribution of oligonucleotides into fat. Oligonucleotides are polar hydrophilic molecules and distribute preferentially to lean body mass (Phillips et al., 1997). Dosing on an actual weight basis did not reflect the patients' volume of distribution, which correlates with lean body weight, and resulted in higher plasma levels compared with lean subjects receiving the same 2 mg/kg dose. For more reproducible plasma levels, it may be appropriate to dose by ideal body weight or with a fixed ISIS 104838 dose.

Following s.c. injection, the apparent distribution half-life is somewhat slower than with i.v. dosing due to the continued absorption from the injection site during the distribution phase. The plasma bioavailability of s.c. injected oligonucleotides has been shown to underestimate the ultimate absorption and tissue availability in monkey studies (Leeds et al., 2000). The most likely explanation for the discrepancy between plasma and tissue bioavailability estimates relates to the rapid distribution from plasma into tissue, coupled with slow absorption from the injection site. Additionally, a portion of the absorption from the s.c. injection site in monkeys is via lymphatic uptake, evidenced by high drug concentrations accumulating in draining lymph nodes. Last, limitations in assay sensitivity have previously prevented full evaluation of the plasma concentration curve following s.c. injection. The hybridization ELISA method measuring ISIS 104838 in plasma for this study was two orders of magnitude more sensitive than the previous CGE assay. Regardless, the plasma bioavailability reported here in human volunteers may still underestimate the total tissue absorption of s.c. injected ISIS 104838, particularly for the lower doses. Higher milligram per milliliter concentrations of s.c. ISIS 104838, and thus higher doses, result in higher net plasma bioavailability. This phenomenon may be a consequence of a shift from lymphatic to direct absorption into blood due to a saturation of the lymphatic flow-limited uptake process. Additional studies will be required to determine optimal concentration and volumes. Subcutaneous injections at the highest concentration (200 mg/ml) were well tolerated with mild tenderness and less commonly mild induration or erythema.

The lower C_max following a s.c. injection, compared with an equivalent i.v. dose, provided a greater safety margin with regard to potential acute side effects related to transient plasma protein binding. Reversible aPTT prolongation has been previously reported with phosphorothioate oligodeoxynucleotides, due to reversible protein binding of the intrinsic tenase complex, and is not sequence-specific (Sheehan and Lan, 1998; Sheehan and Phan, 2001). The amount of prolongation is similar between a phosphorothioate oligodeoxynucleotide i.v. and this MOE gapmer. The mean aPTT prolongation was 12.9 s, or a 56% increase over pretreatment, following a 2-h infusion of the first generation phosphorothioate oligonucleotide ISIS 2302 at 2 mg/kg (Glover et al., 1997). This compared with 10.7 s aPTT prolongation, or 36% increase over pretreatment, following a 1 h infusion of ISIS 104838 at 2 mg/kg. These findings appear to be attributable to the overall chemical class and not to the specific antisense target. The transient aPTT change did not compromise clinical hemostasis and was minimal following s.c. injections. At high doses, particularly in preclinical studies, complement activation has also been seen due to binding of complement Factor H (Henry et al., 1997, 2002). There was no evidence for complement activation in this trial following normal volunteer treatment at a maximum MOE gapmer dose of 6 mg/kg.

Rashes were observed following i.v. administration but were not definitively linked to ISIS 104838, because there was no evidence for either cytokine release or IgE-mediated hypersensitivity. The transient change in platelet count was minor (191 to 113 × 10⁹/liter) and resolved the next day, suggesting sequestration. Previous experience with first generation oligonucleotides in rheumatoid arthritis patients showed treatment to be well tolerated (Maksymowych et al., 2002), and this initial evaluation of a TNF- antisense oligonucleotide will permit future testing of ISIS 104838 in active rheumatoid arthritis patients.