Khachaturian ZS, Lombardo J.
In silico modeling system: a national research resource for simulation of complex brain disorders.
Alzheimers Dement. 2009 Jan;5(1):1-4.
PubMed.
The position paper by Khachaturian and Lombardo is a timely invitation to mobilize and leverage existing expertise in the field of computational neurology to address the burning issue of large failure rates of CNS drugs in clinical development.
The authors correctly identify some of the major problems with the current drug discovery paradigm, amongst them the poor predictability of animal models, which are at best able to capture only part of the complete and complex human pathology. We would like to add 1) some neuronal circuits in the human brain are wired differently than in the rodent, 2) metabolism of drugs is different in rodents, making it unlikely to achieve similar exposures as in the human situation, 3) not all clinically relevant functional human genotypes, such as COMT, can be replicated in rodents, 4) the fact that drugs have often different affinities for rodent targets compared to the same human targets and 5) fundamental clinical readouts, such as episodic memory, do not exist in rodent models (Geerts, 2009).
The authors rightfully acknowledge the issue of emergent systems, which are a consequence of many interacting networks that are best tackled with complex simulation paradigms because the human brain is notoriously bad at quantitatively assessing the effect of more than two interacting targets. In many circumstances, CNS drugs fail for efficacy because of unanticipated off-target effects which tend to reduce the efficacy of the drug at their primary target, because the effect of the investigative drug interferes with the allowed comedication (note the unusual clinical effect of AChE-inhibitors), or functional genotypes are inadvertently segregated into specific treatment arms. Much of this information is available publicly; we just miss the technology to assess the importance of these issues in clinical development and that is where computational neuropharmacology modeling can play an important role.
It is time to mobilize all the forces and expertise of computational neuroscience to make this tool much more useful for drug discovery, similarly to the success systems biology has enjoyed (De Schutter, 2008).
Systems biology is much better appreciated as an essential tool in pharmaceutical R&D, and many companies have capitalized on this and are offering different tools. With only 8 percent of CNS drugs that enter clinical development successfully reaching the market (Kola and Landis, 2004), there is an urgent need for the pharma industry to redesign and re-engineer the drug discovery process. Similarly to other successful industries, such as aerospace and microelectronics, it is in their benefit to rethink their discovery processes to include more mechanistic modeling and simulation as additional decision systems for go-nogo decisions. Having a nationally based resource model that increasingly develops and validates this technology is a first step to demonstrate its commercial usefulness.
References:
Geerts H.
Of mice and men: bridging the translational disconnect in CNS drug discovery.
CNS Drugs. 2009 Nov;23(11):915-26.
PubMed.
Kola I, Landis J.
Can the pharmaceutical industry reduce attrition rates?.
Nat Rev Drug Discov. 2004 Aug;3(8):711-5.
PubMed.
De Schutter E.
Why are computational neuroscience and systems biology so separate?.
PLoS Comput Biol. 2008 May;4(5):e1000078.
PubMed.
Comments
Certara
The position paper by Khachaturian and Lombardo is a timely invitation to mobilize and leverage existing expertise in the field of computational neurology to address the burning issue of large failure rates of CNS drugs in clinical development.
The authors correctly identify some of the major problems with the current drug discovery paradigm, amongst them the poor predictability of animal models, which are at best able to capture only part of the complete and complex human pathology. We would like to add 1) some neuronal circuits in the human brain are wired differently than in the rodent, 2) metabolism of drugs is different in rodents, making it unlikely to achieve similar exposures as in the human situation, 3) not all clinically relevant functional human genotypes, such as COMT, can be replicated in rodents, 4) the fact that drugs have often different affinities for rodent targets compared to the same human targets and 5) fundamental clinical readouts, such as episodic memory, do not exist in rodent models (Geerts, 2009).
The authors rightfully acknowledge the issue of emergent systems, which are a consequence of many interacting networks that are best tackled with complex simulation paradigms because the human brain is notoriously bad at quantitatively assessing the effect of more than two interacting targets. In many circumstances, CNS drugs fail for efficacy because of unanticipated off-target effects which tend to reduce the efficacy of the drug at their primary target, because the effect of the investigative drug interferes with the allowed comedication (note the unusual clinical effect of AChE-inhibitors), or functional genotypes are inadvertently segregated into specific treatment arms. Much of this information is available publicly; we just miss the technology to assess the importance of these issues in clinical development and that is where computational neuropharmacology modeling can play an important role.
It is time to mobilize all the forces and expertise of computational neuroscience to make this tool much more useful for drug discovery, similarly to the success systems biology has enjoyed (De Schutter, 2008).
Systems biology is much better appreciated as an essential tool in pharmaceutical R&D, and many companies have capitalized on this and are offering different tools. With only 8 percent of CNS drugs that enter clinical development successfully reaching the market (Kola and Landis, 2004), there is an urgent need for the pharma industry to redesign and re-engineer the drug discovery process. Similarly to other successful industries, such as aerospace and microelectronics, it is in their benefit to rethink their discovery processes to include more mechanistic modeling and simulation as additional decision systems for go-nogo decisions. Having a nationally based resource model that increasingly develops and validates this technology is a first step to demonstrate its commercial usefulness.
References:
Geerts H. Of mice and men: bridging the translational disconnect in CNS drug discovery. CNS Drugs. 2009 Nov;23(11):915-26. PubMed.
Kola I, Landis J. Can the pharmaceutical industry reduce attrition rates?. Nat Rev Drug Discov. 2004 Aug;3(8):711-5. PubMed.
De Schutter E. Why are computational neuroscience and systems biology so separate?. PLoS Comput Biol. 2008 May;4(5):e1000078. PubMed.
Make a Comment
To make a comment you must login or register.