ICYMI – the New England Patriots won the Super Bowl. How they did it was remarkable, and improbable. To introduce this week’s articles on human genetics and drug discovery, I want to focus on the interception of Russell Wilson by Malcolm Butler. If the pass is on-target, Seahawks win. By now you know the story: the pass was off-target, and the Seahawks lost.
As in football, on-target vs off-target events are highly relevant in drug discovery. Think about what it takes to develop a drug, and how “drug accuracy” (like passing accuracy) can make-or-break a development program. First, you start with a target. Next, you develop a drug against that target. Then, you test the target in pre-clinical models to make sure it is doing what you think it should do. And finally, you take the drug into humans to see if it has an adequate therapeutic index (i.e., is safe and effective).
All along the way you assess whether the therapeutic molecule is selectively engaging and modulating the desired target, and not acting more promiscuously on other targets in the system. Early on, you may use cellular readouts and comparative drugs to assess selectivity. As you progress towards the clinic, pre-clinical animal models are often used. Finally, you assess the effect of the therapeutic molecule in humans.
When you see an “adverse event” (e.g., rash, cough, low blood count), what do you do? One of the first questions you address is whether the adverse event is “on-target” (due to the effect of manipulating the target of interest in the desired way) or “off-target” (due to an effect on something else in the system). The distinction is critical, as the action you take to resolve the adverse event differs based on the two categories…and in the case of pre-clinical models, whether the adverse event is relevant at all in humans.
Two articles in the past week highlight the role of human genetics in evaluating adverse events, especially those that are on-target. Both articles represent works-in-progress as they pertain to drug development, but are nonetheless quite interesting from the perspective of human genetics and adverse drug events. Disclaimer: I am a Merck/MSD employee. The opinions I am expressing are my own and do not necessarily represent the position of my employer.
Effect of selective LRRK2 kinase inhibition on nonhuman primate lung, Fuji et al Science Translational Medicine (February 2015). Mutations in the leucine-rich repeat kinase (LRRK2) gene are the most common genetic cause of both familial and sporadic Parkinson’s disease (PD) (see here, here, here, here). The most common PD-associated LRRK2 mutation, G2019S, has been unequivocally demonstrated to confer increased kinase activity, while a 3-amino acid haplotype with decreased kinase activity is associated with decreased PD risk. These genetic findings are consistent with a dose-response relationship between LRRK2 perturbation and susceptibility to PD. Thus, inhibition of LRRK2’s kinase function is a potential therapeutic target for the treatment of PD. The benefit and risk of inhibiting the kinase activity of LRRK2 is unknown and is currently untested in humans.
Using two selective LRRK2 kinase inhibitors, Fuji et al. report a safety liability in nonhuman primates characterized by morphological changes in lung. These results are consistent with observations in mice lacking LRRK2. These safety observations offer a cautionary note for pharmacological modulation of LRRK2 in humans. The authors argue that the lung pathology is due to an on-target effect of LRRK2 mutation. As the authors conclude: “Our results suggest that lung toxicity may be the primary clinical safety liability of LRRK2 kinase inhibitors in patients.”
How can human genetics help assess on-target adverse events? One intriguing observation is that no pulmonary defects have been observed in carriers of LRRK2 loss-of-function mutations. This provides some support that lifelong inhibition of LRRK2 – at least to the level sufficient for some protection against PD – does not lead to overt lung disease. Thus, a key area for future research will be around the therapeutic window of LRRK2 inhibition. How much LRRK2 inhibition is needed to be beneficial in the brain (protecting from PD) without undesired effects in the lungs?
New loci and coding variants confer risk for age-related macular degeneration in East Asians, Cheng et al Nature Communications (February 2015). Human genetics has identified CETP as a gene important in influencing both HDL and LDL levels, raising the possibility that it is a therapeutic target for coronary heart disease (CHD; see balanced discussion by Dr. Sekar Kathiresan here). In the Nature Communications study by Cheng et al, the authors identified a low-frequency East Asian-specific mutation in CETP (D442G), which they found to be associated with exudative age-related macular degeneration (AMD, per-allele OR=1.70). Supporting the importance of CETP in the pathogenesis of AMD, previous studies have shown that a common variant (rs3764261) mapping to the intergenic region between HERPUD1 and CETP is associated with AMD in Europeans (OR=1.15). Molecular studies have shown that the mutant 442G allele impairs CETP function with reduction in plasma CETP mass and activity. Importantly, the allele that increases risk of AMD is known to protect from CHD.
How can human genetics help predict on-target adverse drug events related to CETP modulation? The Nature Communications study suggests that modulation of CETP in the direction that lowers LDL and raises HDL may increase risk of AMD. The available clinical trial data do not yet support this prediction, however. As with the discussion above (LRRK2 and lung pathology), the answer could lie in the therapeutic index. Along these lines, the effects on risk of AMD relate to lifelong perturbation of CETP (via human genetics), whereas therapeutic modulation occurs for a shorter duration of time (albeit still for many years in the case of CHD protection).
All drugs have adverse events. This in and of itself should not stop drug development. Still, it is important to use all available tools thoroughly assess efficacy and safety. Human genetics represents one powerful tool that should be used to assess therapeutic index, especially those adverse events that are due to on-target effects. These embedded genetic tools will increasingly become available, as should be clear from efforts such as that announced recently by the US at the White House (Precision Medicine Initiative).