If you share my penchant for investing in biotech and pharma, then you will have, no doubt, come across the standard analyst practice of evaluating the NPV of a pipeline. This is critical but, how is it done? Moreover, does the potential blockbuster in your pharma co's pipeline really have as strong a chance of succeeding as you think it does?
The usual method is to evaluate the pipeline in terms of the potential end market of the indication, market share, and then the probability of the success rate of each individual program. I want to focus on the last factor in this calculation, even though the whole process is fraught with uncertainty. Specifically, I want to discuss probabilities of success in clinical trials. I also want to highlight how varied these rates can be and why?
Big Pharma vs. Small Pharma
The usual approach taken by analysts is to assume a set of probabilities of success for each stage of the trial. How many times have you come across an analysis of a late stage drug, whereby the analyst has penciled in a 50% chance of success for the trial? I’ve seen this umpteen times, but rarely are these analyses accurate.
I suspect this is because the rates are taken from industry averages, which do not explain many factors which go toward their creation. An example of a typical probability analysis, for a listed pharma company, is shown in the top line of this table.
So, for example, you might see a drug in Phase II and the analyst has penciled in (.57*.57*.8) a probability of success of 25% or, a Phase III drug might be given a (.57*.8) 46% chance of being marketed.
The first thing to note is that big pharma is much better in Phase III. The chances are 55% vs. small pharma chances of 43%. I think the reasons for this are that late stage trials are large and expensive. They require resources and, they require experience. In addition, many small pharma companies partner with big pharma for later stage trials and, for good reason.
Takeaway: Favour big pharma over small in late stage and always ask yourself why your small pharma company hasn’t signed a partnership deal.
As an example, YM Biosciences (AMEX: YMI) is a good example with its lead compound CYT387. This is a Janus Kinease (JAK) inhibitor which is being developed to treat myelofibrosis (a bone marrow disorder). In early stage trials, it has demonstrated the ability to reduce spleen size and –unlike Incyte’s (NASDAQ: INCY) Jakafi- has also demonstrated that it could reduce anemia.
However, despite rumors of discussions, it hasn’t been partnered with big pharma yet. All of which, leaves a small company trying to get through Phase III on its own. I would suggest to adjust the probability of success lower as a result.
Molecule Type Matters
Borrowing from a separate analysis, I want to look at how molecule type plays a part.
New Molecular Entity (NME) just means that the molecule contains no active ingredient that has been approved by the FDA for any other application. It is clear that novel drugs and/or modes of action have far less success rates.
Moreover, non-NMEs have quite high rates. It is no wonder that the pharma companies are criticized for releasing ‘me-too’ drugs. They do it for a reason. The drugs work. Also, note that lead indications are typically 3-5x more likely to succeed than secondary indications.
As for biologics, they tend to treat rarer diseases so their primary endpoints can be less stringent and they may have less competition. Therefore establishing efficacy over an existing drug is not so much of an issue.
Takeaways: ‘Me-too’ drugs within an established class and mode of action, have far higher chances of success than novel drugs do. If a molecule’s lead indication data is weak, don’t hold out for hope going for a secondary indication.
Examples: Vertex (NASDAQ: VRTX) had a protease inhibitor (Incivek) approved for Hepatitis C treatment by the FDA. On the same day, Merck also got a protease inhibitor (Victrelis) approved for the same indication. Moreover, Bristol Myers, Gilead(NASDAQ: GILD), Boehringer Ingelheim and others, also have protease inhibitors in development. I would suggest that the chances of getting these drugs approved, are relatively high.
In addition, I’ve written about JAK inhibitors in a detailed article linked here. Incyte’s Jakafi (myelofibrosis) was the first approved in this class, but now Pfizer (NYSE: PFE) has a potential blockbuster with tafocitinib set for approval. Similarly, Vertex has VX509 in development for Rheumatoid Arthritis as does Incyte with INCB28050. In particular, Incyte and Vertex seem to have a number of drugs in development that suggest their chances of success are relatively high.
The Type of Disease Matters, It Matters a Lot
Tabulating rates for some selected diseases.
It is apparent that the chances of getting a drug marketed does, superficially at least, appear to vary depending on the target indication.
However, what this data doesn't do is breakdown the proportion of NMEs to non-NMEs within it. In addition, it doesn’t analyze the value of the programs. It stands to reason that drugs with higher market potential will receive higher research and trial funding. Moreover, an indication like AIDS maybe granted accelerated approval process or less stringent criteria, due to its necessity. In addition, The FDA does factor in the availability of other drugs when giving feedback over the planning of late stage trials.
That said, there are some striking differences here. In particular, note the lack of success with Alzheimer's.
Takeaways: Understand the nature of the target disease and think about why molecules have failed. Try to look into the clinical need for drug, as it does influence the end point, safety and efficacy selection.
Examples: I’ve previously written about a class of drugs which rely on the amyloid-beta theory to treat Alzheimer’s in an article linked here and, given their lack of previous clinical success I think it is safe to assume a relatively lower probability.
Alzheimer’s is a condition which has seen relatively meager breakthroughs in medical understanding in the last decade and, I would not give Pfizer’s bapineuzumab or Eli Lilly’s solanezumab a high chance of success. However, it may not matter, because the value of a new treatment in this indication will be very high. So drugs companies will pursue treatments even with a low chance of success.
On another note, consider AIDs/HIV medications, where Gilead is the clear leader. Its franchise was built on one initial antiviral called Viread, which goes off patent in 2017. The later drugs in the franchise (Truvada, Atripla, Quad) are all combination drugs which have Viread as a base. I think it is fair to argue, that the high degree of success with AIDS drugs is partly due to the submission of non-NMEs that expand the franchise and because existing treatments were lacking.
Oncology a Special Case?
I want to look closer at cancer because within this disease, there are vastly different success rates. The following data is from the Biotechnology Industry Organization and includes many small and private company data, so don’t worry if it looks incongruent with the data above. The key is the relative success rates.
Conclusion
In putting all these elements together, I will share an anecdote over a UK pharma company, namely Antisoma.
A few years ago, I looked at it. It had a pipeline of oncology products. Analysts were very bullish about its prospects with its lead compound, an NME in a class of oncology drugs called vascular disrupting agents. It had given excellent results in a Phase II trial in NSCLC but not so sterling results in ovarian or prostate cancer. I was initially enthusiastic. I read analyst reports which talked of a potential blockbuster and the chances of approval ranged from 30-40%!
So, in summary, we have..
It now trades at 1.7p
Investors should take the practice of analysts mechanically penciling in probabilities of success, with a pinch of salt. In reality, there are many factors which will give a better 'guesstimate' to these numbers, than just merely assuming a drug has a 50% ‘pop’ at passing phase III.
I don’t pretend to possess the answer to calibrating these numbers, but I hope this article is useful in shedding light on the problem!
Source:
Rosa M.Abrantes-Metz, Christopher P. Adams, Albert Metz "Pharmaceutical Development Phases: A Duration Analysis" Federal Trade Commission Bureau of Economics, 2004
The usual method is to evaluate the pipeline in terms of the potential end market of the indication, market share, and then the probability of the success rate of each individual program. I want to focus on the last factor in this calculation, even though the whole process is fraught with uncertainty. Specifically, I want to discuss probabilities of success in clinical trials. I also want to highlight how varied these rates can be and why?
Big Pharma vs. Small Pharma
The usual approach taken by analysts is to assume a set of probabilities of success for each stage of the trial. How many times have you come across an analysis of a late stage drug, whereby the analyst has penciled in a 50% chance of success for the trial? I’ve seen this umpteen times, but rarely are these analyses accurate.
I suspect this is because the rates are taken from industry averages, which do not explain many factors which go toward their creation. An example of a typical probability analysis, for a listed pharma company, is shown in the top line of this table.
So, for example, you might see a drug in Phase II and the analyst has penciled in (.57*.57*.8) a probability of success of 25% or, a Phase III drug might be given a (.57*.8) 46% chance of being marketed.
The first thing to note is that big pharma is much better in Phase III. The chances are 55% vs. small pharma chances of 43%. I think the reasons for this are that late stage trials are large and expensive. They require resources and, they require experience. In addition, many small pharma companies partner with big pharma for later stage trials and, for good reason.
Takeaway: Favour big pharma over small in late stage and always ask yourself why your small pharma company hasn’t signed a partnership deal.
As an example, YM Biosciences (AMEX: YMI) is a good example with its lead compound CYT387. This is a Janus Kinease (JAK) inhibitor which is being developed to treat myelofibrosis (a bone marrow disorder). In early stage trials, it has demonstrated the ability to reduce spleen size and –unlike Incyte’s (NASDAQ: INCY) Jakafi- has also demonstrated that it could reduce anemia.
However, despite rumors of discussions, it hasn’t been partnered with big pharma yet. All of which, leaves a small company trying to get through Phase III on its own. I would suggest to adjust the probability of success lower as a result.
Molecule Type Matters
Borrowing from a separate analysis, I want to look at how molecule type plays a part.
New Molecular Entity (NME) just means that the molecule contains no active ingredient that has been approved by the FDA for any other application. It is clear that novel drugs and/or modes of action have far less success rates.
Moreover, non-NMEs have quite high rates. It is no wonder that the pharma companies are criticized for releasing ‘me-too’ drugs. They do it for a reason. The drugs work. Also, note that lead indications are typically 3-5x more likely to succeed than secondary indications.
As for biologics, they tend to treat rarer diseases so their primary endpoints can be less stringent and they may have less competition. Therefore establishing efficacy over an existing drug is not so much of an issue.
Takeaways: ‘Me-too’ drugs within an established class and mode of action, have far higher chances of success than novel drugs do. If a molecule’s lead indication data is weak, don’t hold out for hope going for a secondary indication.
Examples: Vertex (NASDAQ: VRTX) had a protease inhibitor (Incivek) approved for Hepatitis C treatment by the FDA. On the same day, Merck also got a protease inhibitor (Victrelis) approved for the same indication. Moreover, Bristol Myers, Gilead(NASDAQ: GILD), Boehringer Ingelheim and others, also have protease inhibitors in development. I would suggest that the chances of getting these drugs approved, are relatively high.
In addition, I’ve written about JAK inhibitors in a detailed article linked here. Incyte’s Jakafi (myelofibrosis) was the first approved in this class, but now Pfizer (NYSE: PFE) has a potential blockbuster with tafocitinib set for approval. Similarly, Vertex has VX509 in development for Rheumatoid Arthritis as does Incyte with INCB28050. In particular, Incyte and Vertex seem to have a number of drugs in development that suggest their chances of success are relatively high.
The Type of Disease Matters, It Matters a Lot
Tabulating rates for some selected diseases.
It is apparent that the chances of getting a drug marketed does, superficially at least, appear to vary depending on the target indication.
However, what this data doesn't do is breakdown the proportion of NMEs to non-NMEs within it. In addition, it doesn’t analyze the value of the programs. It stands to reason that drugs with higher market potential will receive higher research and trial funding. Moreover, an indication like AIDS maybe granted accelerated approval process or less stringent criteria, due to its necessity. In addition, The FDA does factor in the availability of other drugs when giving feedback over the planning of late stage trials.
That said, there are some striking differences here. In particular, note the lack of success with Alzheimer's.
Takeaways: Understand the nature of the target disease and think about why molecules have failed. Try to look into the clinical need for drug, as it does influence the end point, safety and efficacy selection.
Examples: I’ve previously written about a class of drugs which rely on the amyloid-beta theory to treat Alzheimer’s in an article linked here and, given their lack of previous clinical success I think it is safe to assume a relatively lower probability.
Alzheimer’s is a condition which has seen relatively meager breakthroughs in medical understanding in the last decade and, I would not give Pfizer’s bapineuzumab or Eli Lilly’s solanezumab a high chance of success. However, it may not matter, because the value of a new treatment in this indication will be very high. So drugs companies will pursue treatments even with a low chance of success.
On another note, consider AIDs/HIV medications, where Gilead is the clear leader. Its franchise was built on one initial antiviral called Viread, which goes off patent in 2017. The later drugs in the franchise (Truvada, Atripla, Quad) are all combination drugs which have Viread as a base. I think it is fair to argue, that the high degree of success with AIDS drugs is partly due to the submission of non-NMEs that expand the franchise and because existing treatments were lacking.
Oncology a Special Case?
I want to look closer at cancer because within this disease, there are vastly different success rates. The following data is from the Biotechnology Industry Organization and includes many small and private company data, so don’t worry if it looks incongruent with the data above. The key is the relative success rates.
Conclusion
In putting all these elements together, I will share an anecdote over a UK pharma company, namely Antisoma.
A few years ago, I looked at it. It had a pipeline of oncology products. Analysts were very bullish about its prospects with its lead compound, an NME in a class of oncology drugs called vascular disrupting agents. It had given excellent results in a Phase II trial in NSCLC but not so sterling results in ovarian or prostate cancer. I was initially enthusiastic. I read analyst reports which talked of a potential blockbuster and the chances of approval ranged from 30-40%!
So, in summary, we have..
- A small company (although in partnership with Novartis) developing a drug in Phase III
- An NME
- A difficult disease target (cancer)
- A lead compound with good results in a very tough indication (NSCLC) within oncology, but not so good in an easy indication
It now trades at 1.7p
Investors should take the practice of analysts mechanically penciling in probabilities of success, with a pinch of salt. In reality, there are many factors which will give a better 'guesstimate' to these numbers, than just merely assuming a drug has a 50% ‘pop’ at passing phase III.
I don’t pretend to possess the answer to calibrating these numbers, but I hope this article is useful in shedding light on the problem!
Source:
Rosa M.Abrantes-Metz, Christopher P. Adams, Albert Metz "Pharmaceutical Development Phases: A Duration Analysis" Federal Trade Commission Bureau of Economics, 2004
No comments:
Post a Comment