The Case Of The Malodorous Metabolite
Over where I also write at Forbes.com, Matthew Herper, Senior Editor for Pharma and Healthcare, reported on GlaxoSmithKline’s NEJM clinical trial of darapladib in coronary artery disease. The drug had been developed as a small molecule, orally-formulated, subnanomolar inhibitor of lipoprotein-associated phospholipase A2.
Lp-PLA2 is associated with apolipoprotein B-containing lipoproteins, such as LDL and non-HDL particles, and is found in the necrotic center of atherosclerotic plaques. Formerly known as platlet-activating factor acetylhydrolase, the enzyme produces pro-apoptotic lysophosphatidylcholine and stimulates the synthesis and release of proinflammatory mediators such as IL-1 beta, IL-6, ICAM-1, and VCAM-1.
A meta-analysis of nearly 80,000 patients across 32 trials indicated that Lp-PLA2 was associated with a relative risk for coronary artery disease of 1.10 for each 1-standard deviation increase in activity, even when correcting for other influences. This magnitude of relative risk is similar to that for non-HDL cholesterol or systolic blood pressure.So, the enzyme seemed like a reasonable target for small-molecule inhibition.
Darapladib (SB-480848) emerged from screening for inhibitors and a synthesis campaign, each published in 2002 and 2003, and was selected for clinical trials. A study with 330 patients with coronary artery disease, The Integrated Biomarkers and Imaging Study-2 trial, showed that 12 months of darapladib (160 mg daily, p.o.) decreased the progression of atherosclerotic plaques that occurred in the placebo group, even when they were also receiving standard-of-care statin therapy.Last week saw the publication in NEJM of the STABILITY trial to examine efficacy against primary endpoints. Unfortunately, this large, placebo-controlled, randomized trial with 15,828 patients who took 160 mg darapladib once daily showed no benefit of the drug relative to placebo in the primary endpoints of time to cardiovascular death, myocardial infarction, and stroke. However, the drug did show significant differences from placebo in measures of major coronary events (9.3% vs. 10.3 % for placebo) and total coronary events (14.6% vs. 16.1% for placebo).
Major coronary events were defined as “a composite of death from coronary heart disease, myocardial infarction, or urgent coronary revascularization for myocardial ischemia.” Total coronary events were defined as “a composite of death from coronary heart disease, myocardial infarction, hospitalization for unstable angina, or any coronary revascularization procedure.”
But when looking at other aspects of the paper, Herper noted that a small but significant subset of patients reported an unpleasant odor from their skin, urine, or feces while taking darapladib. In discussing the side effects relative to the modest effects on secondary endpoints, Herper wrote,
[T]he leading side effects were diarrhea (3.2% vs. 0.8%), feces odor (2.2% vs. 0.1%), urine odor (1.4% vs. <0.1%), and skin odor (2.2% vs. 0.1%).
Would such a small difference be enough to lead cardiologists to prescribe a drug that will leave them fielding tons of calls from patients asking, “Doc, why do I smell so bad?”
Of course, I read that paper trying to figure out why the drug or its metabolites might smell unpleasant. No discussion was apparent but I appreciate Herper catching this aspect of the trial.
(Note to my ENG 507 health and environmental writing students at NC State: This is yet another good reason to read the entire paper in pulling out story ideas. Science writer Jennifer Ouelette told last year’s class the she is particularly fond of reading Materials and Methods of papers to find quirky approaches.)
So I went to GSK’s in vitro and in vivo darapladib metabolism paper, just published last month. The authors conducted a single dose i.v. study and a single dose and 21-day oral dosing studies.
Only two major CYP-mediated metabolites were detected after oral dosing in human subjects: an N-deethylated metabolite with pharmacological activity (M4, above) and an inactive hydroxylated metabolite (M3). Various gut bacteria incubated under anaerobic conditions failed to produce any significant metabolites.
But the authors also detected an acid hydrolysis product (M10) that liberated 4-fluorobenzyl mercaptan (p-fluorotoluene-α-thiol; CAS 15894-04-9). While they didn’t comment on the smell of this compound, they did note that they ultimately chose to make an enteric-coated darapladib dosage form for clinical trials.
Most readers here would predict this metabolite was the source of the malodor. The relatively low incidence of unpleasant smell from the skin, urine, and feces would be indicative of the general efficacy of the enteric coating and interindividual variations in gastric acidity, emptying time, and patient compliance with the directive to take with food. The metabolite from which 4-fluorobenzyl mercaptan was liberated comprised up to 5% of total metabolites.
Benzyl mercaptan is known to give a smoky odor to wines, but that doesn’t sound particularly unpleasant. It took going to ChemSpider to find any odor characteristics of 4-fluorobenzyl mercaptan. Their citation of Matrix Scientific’s offering of the compound lists hazards as, “TOXIC, STENCH.”
Do any of you have experience with this compound? Would you expect even a small amount of it as a metabolite to be responsible for patient reporting of such an odor?
Looking back at the 2008 IBIS-2 trial, I found this statement from the authors:
A higher incidence of malodor (mainly feces or urine) was reported with darapladib (16%) compared with placebo (3%) but was an uncommon cause of withdrawal from the study (darapladib, 2%).
So, this side effect had been known previously.
Interestingly, the 4-fluorobenzyl mercaptan substituent was on the original compound, SB-435495, that emerged from GSK’s screen prior to lead optimization.
I’m not sure where darapladib will be going after this. But GSK’s press release on the STABILITY study noted that another study was ongoing:
Dr Murray Stewart, Senior Vice President, Metabolic Pathways Cardiovascular Therapy Area, added:
STABILITY was a robust, large-scale cardiovascular outcomes study of a novel mechanism with the goal of providing incremental benefit above a high level of standard of care. Given the unmet medical need, the results of the STABILITY study are important in understanding how this mechanism may impact the lives of patients with heart disease. We await the results of the second study, SOLID-TIMI 52, to better understand the findings.
The second Phase III study, SOLID-TIMI 52 will evaluate the effects of darapladib in patients with acute coronary syndrome. The trial has enrolled over 13,000 patients across 36 countries. SOLID-TIMI 52 is ongoing and remains blinded. Results are expected in the second quarter of 2014.
We’ll see what comes next and what success investigators have made in minimizing the malodorous metabolite.