Search

Aim: To analyze the late-stage failures of monoclonal antibody drugs. The later a drug fails in development, the more time and expense is incurred by the sponsor., Methods: We review the late stage, Phase III, failures of 21 monoclonal antibody drugs between 2014 and 2019 using published and publicly available information to characterize the reasons for these failures., Results: In some cases, the failures are unavoidable due to the lack of adequate science, but in others, we characterize the causes of such failures and recommend how such failures may have been avoided., Conclusion: By learning from previous mistakes and adhering to the principles and recommendations provided, it is possible to avoid these common pitfalls, increasing the likelihood of success in phase III clinical trials, and thus securing regulatory approval., (© 2020 S. Karger AG, Basel.)

Recent studies in our laboratory have shown that the loop diuretic, furosemide, is actively secreted by Caco-2 cells and rat jejunal tissue. This active secretion could be the result of efflux transporters such as P-gp, MRP1 or MRP2 (cMOAT). To determine if any of these transporters is responsible for the secretion of furosemide, we compared directional permeability in the wild-type cell lines, MDCK strains I and II, and LLC-PK1, vs. cell lines that overexpress a single transporter, in both the presence and absence of various inhibitors, for furosemide as compared to vinblastine. Sulfinpyrazone significantly inhibited the transport of vinblastine in MRP2 expressing cells, but not the wild-type controls. Vinblastine could not be confirmed as a substrate of MRP1. We were also unable to demonstrate that any particular transporter affected furosemide in excess of the background effects of endogenous transporters in the parental cell lines. Furosemide secretion from these kidney-derived cell lines is probably not the primary result of any of the well characterized efflux transporters (P-gp, MRP1 or MRP2), although they may still play a role in the observed Caco-2 secretion. This equivocal result acknowledges the difficulty in trying to determine the effect of a single protein in a complicated expression system., (Copyright 2002 S. Karger AG, Basel)

4-Dimethylamino-4'-(imidazol-1-yl)chalcone (RL3142) is a newly developed antimalarial cysteine protease inhibitor. Four metabolites (M1-M4) were found in human liver microsomes and their structures were identified by LC/MS/MS. Two primary metabolites, M2 (minor) and M4 (major), were determined to be the N-demethylated product (M2) and the product (M4) resulting from 1,2-hydrogenation of the alpha, beta-unsaturated ketone moiety of the parent compound. A combined approach utilizing selective P450 inhibitors, immunoinhibition with CYP3A and NADPH P450 reductase antibodies, and cDNA expressed human CYP3A4 and NADPH P450 reductase, was used for identification of enzymes responsible for the biotransformation. For formation of M2, both a rabbit CYP3A polyclonal antibody (110 microliter/mg microsomal protein) and ketoconazole (2 micromol/l), a CYP3A inhibitor, showed about 50% inhibitory effects; other specific inhibitors of CYP1A2, CYP2B6, CYP2C9, CYP2C19, CYP2D6 and CYP2E1 showed no significant effects. For formation of M4, neither CYP3A antibody nor the above mentioned CYP inhibitors exhibited inhibitory effects. Anti-rat NADPH P450 reductase serum (50 microliter/100 microgram microsomal protein) exhibited 70 and 58% inhibitory effects on M2 and M4 formation, respectively. Incubation of RL3142 with cDNA expressed human NADPH P450 reductase yielded formation of M4, but not M2. Carbon monoxide inhibited formation of M2 and M1 (the reduced product of M2), but had no effect on M4 and M3 (the reduced product of M4) formation. Collectively, NADPH P450 reductase solely catalyzed reduction of RL3142 to M4, whereas CYP3A contributed in part to formation of M2.

Digoxin, a cardiac glycoside, is a substrate of the multidrug transporter P-glycoprotein (Pgp), and in rats has also been identified as a substrate for cytochrome P450 3A (CYP3A). Ketoconazole, an antifungal agent, was shown to inhibit Pgp in a multidrug-resistant cell line, and is known to be a potent inhibitor of CYP3A. Here, we determined the effects of ketoconazole on digoxin absorption and disposition in rats. Digoxin was administered intravenously or orally with or without a concomitant oral dose of ketoconazole. When given intravenously, digoxin AUC increased from 93 +/- 22 to 486 +/- 26 microg x h/l with ketoconazole administration. Similarly, ketoconazole raised the AUC of orally administered digoxin from 63 +/- 17 to 411 +/- 50 microg x h/l. Concomitant ketoconazole administration prolonged digoxin elimination, yielding a nonlinear pharmacokinetic profile. Using time-averaged values, digoxin bioavailability increased from 0.68 +/- 0.18 to 0.84 +/- 0.10, while mean absorption time was reduced from 1.1 +/- 0.4 to 0.3 +/- 0.1 h. Thus, in rats, ketoconazole increases digoxin plasma concentrations, rate of absorption and bioavailability. Although the effects of ketoconazole on AUC could be explained by inhibition of both CYP3A and Pgp, which cannot be differentiated in this study, the decreased mean absorption time can only be explained by inhibition of Pgp in the intestine.

Based on previous observations of very short periods of linearity for antiprogestin metabolite formation and the presence of a common tertiary amine moiety in each compound as the principal site of their metabolism, we hypothesized that mifepristone, lilopristone and onapristone are oxidized by cytochrome P450 (CYP) 3A4 to reactive nitroso species that complex the heme of the enzyme, thereby inactivating it. Upon preincubation with human liver microsomes in the presence (but not the absence) of NADPH, mifepristone inhibited midazolam 1'-hydroxylation, a marker of CYP3A4 catalytic activity, very potently (IC50 approximately 3.5 mumol/l) and extensively (by approximately 87%). Lilopristone and onapristone also displayed NADPH and time-dependent inactivation of CYP3A4 with characteristics very similar to mifepristone. These data support antiprogestin-mediated inactivation of CYP3A4 and suggest the potential for drug-drug interactions and time-dependent nonlinearities in pharmacokinetics upon their long-term administration.

In a previous study, the disposition and irreversible plasma protein binding of tolmetin have been described in young volunteers. Significant levels of tolmetin glucuronide and its isomer(s) were found in plasma, and irreversible binding was shown to occur in all subjects. In the present study, the pharmacokinetics and irreversible binding in 5 elderly volunteers are investigated after a single dose (400 mg) of tolmetin. No difference of physiological importance has been found, unless weight correction is employed, when compared with the young subjects. Very strong correlations were present between the level of binding and exposure to the glucuronide or its isomer. An elderly patient currently treated with tolmetin (400 mg b.i.d.) also participated in the study. The results show an accumulation of the irreversible binding to levels 4-17 times higher than after a single dose. Nevertheless, neither toxic nor allergic reactions have been observed in this patient.

Because of the labile nature of acyl glucuronides under physiologic conditions, metabolic rates of formation calculated using traditional methods may be confounded by concomitant rates of degradation. True metabolic formation rates may be approximated by stabilization of the metabolic product or by purifying the metabolite and calculating its degradation rate under similar conditions to that utilized in the formation experiments. This latter degradation rate is used to correct the apparent formation parameter in the absence of inhibitors. The advantages and limitations of each approach are reviewed employing studies to characterize rates of acyl glucuronidation for nonsteroidal antiinflammatory drugs.

The interconversion and elimination of prednisone (PO) and prednisolone (POH) were examined in human and rabbit liver, kidney, lung and skeletal muscle preparations to determine the effect of organ disruption on in vitro metabolism within multiple organs of two species. Results from microsomes, homogenates and minces were compared to determine the effect of various stages of tissue disruption on the reversible and irreversible metabolism of the glucocorticoids. Oxidation of POH to PO was enhanced by homogenization of all the organs examined; further disruption by subcellular fractionation to microsomes reduced oxidation relative to homogenates but yielded values greater than the original minces. The reverse reaction, reduction of PO to POH, was also enhanced by homogenization, but microsomal preparations were less active than the minces. The irreversible elimination of the glucocorticoids was progressively diminished by disruption of normal architecture in all organs. Results of in vitro studies of glucocorticoid metabolism have provided discrepant results, as a function of the laboratory, the species, the organ and the organ preparation examined. These results provide insights into the source of discrepant results regarding steroid metabolism, as it appears that results of experimentation with glucocorticoid conversion and/or elimination may be a function of the method used. Additionally, the data support the hypothesis that the enzyme system involved in the interconversion of the glucocorticoids may not be a simple single protein which acts bidirectionally.

The diastereomeric glucuronic acid conjugates are major metabolites of the nonsteroidal anti-inflammatory drug fenoprofen (FEN). Glucuronidation of FEN enantiomers was investigated with liver microsomal preparations from different species (sheep, rabbit, rat and human). The formed R- and S-FEN conjugates can be separated and quantitated directly on a C18 reversed-phase HPLC column using a mixture of acetonitrile and tetrabutylammonium sulfate buffer, pH 2.5, as mobile phase. Applying this analytical procedure, it is possible to characterize enantioselective glucuronidation of FEN. For in vitro procedures, rates of glucuronide formation are substrate (FEN) and cosubstrate (UDP glucuronic acid, UDPGA) dependent with initial rates of glucuronide formation being higher for R- than for S-FEN. The R/S ratio of the formed products was independent of UDPGA (2.5-15 mmol/l) and substrate concentrations greater than or equal to 0.4 mmol/l. Enantioselective cleavage of the formed FEN conjugates by alkaline hydrolysis and hydrolytic enzymes (R greater than S-glucuronide) can be controlled during in vitro studies by pH adjustment and the addition of enzyme inhibitors.

The disposition of acetaminophen was studied in adult nonpregnant ewes, neonatal lambs, pregnant ewes and their fetal lambs following single intravenous bolus doses. Comparisons of the pharmacokinetic parameters of acetaminophen were made between sheep at different developmental periods to examine the developmental alteration in the physiological and biochemical processes governing drug disposition. The four groups of animals exhibited distinct and significant differences in the total clearance and metabolic clearances of acetaminophen and the renal clearances of the glucuronide and sulfate conjugates. Placental transfer of acetaminophen from fetus to mother serves as the dominant route of fetal removal or clearance of the drug.

14 infusion studies to steady state with probenecid were carried out in 2 monkeys. Plasma clearance decreased with increasing rate of infusion. The urinary excretion data showed a decrease in the fraction of the dose excreted as probenecid conjugate, an increase in the total N-depropyl metabolites, and no change in the 2-hydroxy metabolites with increasing probenecid dose. Pooled Michaelis-Menten constants for probenecid metabolism were estimated from a plot of the linearized Michaelis-Menten equation using infusion rates of probenecid versus the plasma clearances of the drug.

Since patients vary in their response to furosemide, a pharmacokinetic study of this drug was undertaken in 7 patients with severe congestive heart failure. The extent of oral availability, determined by the ratio of areas under the oral to intravenous plasma concentration-time curve, averaged 61%, but varied widely from patient to patient, from 34 to 80%. Absorption of the drug was often delayed, with a lag time before appearance in the plasma averaging 28 +/- (SD) 15 min. The volume of distribution of furosemide (averaging 114 ml/kg) was not diminished in the presence of congestive failure; likewise, drug clearance and elimination half-lives are similar to values reported in normal subjects. Urinary flow rates 1--2 h following dosing were significantly less after oral than intravenous dosing, but flow rates over 4, 8 or 24 h after dosing were similar for both routes of administration.

The transport of organic anions by the kidney has been shown to be a carrier-mediated process. In an effort to learn more about this process, and examine the potential for two organic anions to compete for the same carrier site, studies were done which involved the transport of p-aminohippuric acid (PAH) and furosemide by vesicles made from basal-lateral membranes of rabbit kidney proximal tubules. Basal-lateral membranes were prepared by differential and ultracentrifugation. The transport was measured by using radiolabelled (3H) organic anions. The transport of each molecule was inhibited by probenecid, indicating that the carrier-mediated process for organic anion transport was functional in these studies. The results indicate that transport of PAH can be inhibited by furosemide in a concentration-dependent manner. This may indicate competition for the same carrier site. Inhibition of furosemide transport by PAH was not significant, perhaps due to much variability in the data. This variability may be due to nonspecific binding of furosemide to the vesicle, higher affinity of furosemide than of PAH for the receptor, or to the presence of more transport carriers for furosemide than for PAH. Experiments were done to determine the extent of nonspecific binding of furosemide. The results show that nonspecific binding of furosemide is extensive, indicating that this may contribute to the differences seen in the inhibition of transport. The data suggest that PAH and furosemide are transported by a common carrier-mediated process in the proximal tubule of the rabbit kidney.

beta-Adrenergic agonist analogs (congeners) of isoproterenol in which the N-isopropyl group has been linked to a p-methyl- (119) or p-trifluoromethyl- (143) anilide moiety through a four carbon methylene spacer have been investigated with respect to their plasma pharmacokinetic profiles and biliary and urinary elimination characteristics in rats. In spite of the differences in selectivity of pharmacologic effects and durations of action between these unique beta-adrenergic agonists and isoproterenol, no differences were observed in their pharmacokinetic parameters in plasma after intravenous administration. Plasma clearances were rapid (67-78 ml/min) and the compounds were widely distributed. In contrast to the known elimination characteristics of isoproterenol, biliary excretion was the major pathway for elimination of 119 and 143. Parent drug and 'one' major metabolite peak appeared in HPLC chromatograms of bile collected from rats that received 119 and 143 by intravenous administration. Preliminary evidence suggests that this metabolite peak consists of one or more glucuronide and/or sulfate conjugates. Urinary excretion appears to be of lesser quantitative importance for 119 and 143 than for isoproterenol. The protracted duration of residence of the derivatives in the heart may help to explain the unusual effects and tissue-specific pharmacological properties of these unique beta-adrenergic agonists.

The purpose of the present study was to determine if furosemide's active transport process could be saturated at therapeutic concentrations and to define a relationship between furosemide in a measurable sampling compartment and its diuretic effect. The experiments utilized Sprague-Dawley rats, ranging in weight from 248 to 313 g, anesthetized with sodium pentobarbital (60 mg/kg). The femoral artery and vein as well as the bladder were cannulated, and samples were taken to measure inulin and furosemide concentrations. 28 rats were infused, after a suitable loading dose (0.5--1.5 mg/kg), to steady-state plasma furosemide levels over the therapeutic concentration range 0.8--25.1 micrograms/ml. Total renal clearance (corrected for kidney function as measured by inulin clearance) showed a negative correlation with plasma concentration (r = -0.655 p less than 0.001), and a good correlation was found between urine flow rate and the urinary excretion rate of furosemide (r = 0.777, p less than 0.001). Steady-state plasma levels of furosemide showed a poor correlation with urine flow rate (r = 0.377, p greater than 0.10).

14 bolus studies with probenecid in a dose range of 8.2--164 mg/kg were carried out in 3 monkeys. The protein binding of probenecid in the plasma was studied by an ultracentrifugation method. Plasma clearances of both total and unbound probenecid decreased with increasing doses, while the volume of distribution and fraction of metabolites excreted in the urine remained constant. The urinary excretion rates of probenecid and it metabolites were measured in a bladder-catheterized monkey and were found to be urine flow dependent. Since the rate-limiting step for the excretion of the metabolites did not depend on the formation rate only, biliary recycling was speculated.

A microanalytical procedure was developed using a gas chromatographic-electron capture technique which is capable of detecting 2 ng pseudoephedrine in 20 microliter plasma samples. Standard curves for pseudoephedrine are linear over the concentration range 0.1--1.6 microgram/ml. Plasma and urine concentrations of pseudoephedrine were followed in 3 rats after intravenous dosing. Derived pharmacokinetics parameters exhibited little inter-animal variation. Average plasma clearance was 67.6 ml/min/kg, with renal clearance averaging 30.3 ml/min/kg. This latter value is approximately 4X the glomerular filtration rate in the rat.

In 6 patients with severe left-sided cardiac failure the absorption of drug from digoxin tablets was assessed by simultaneous administration of a commercially available tablet and of an intravenous tracer dose before and during therapy. Lag time, peak time, and dose-normalized increase in plasma concentration did not change in spite of clinical improvement. Two independent measures of the extent of availability, based on plasma and urine measurements, also remained unchanged and their average was similar to that reported for normals. There was, however, a great interindividual variability in absorption. We conclude that the absorption of digoxin tablets is unchanged in patients with left-sided cardiac failure.