setting the record straight

Mesoblast Ltd (MSB)
Implied Return 94%
Marc Sinatra
+61 3 9200 7050
[email protected]
Setting the record straight
Company Data
ASX Code
Shares on issue
Market capitalisation
12 month price range
ASX turnover (monthly rolling)
MSB
284,478,361
$1.9b
$5.44 - $10.04
16.7m
Board of Directors
Brian Jamieson
Silviu Itescu
Michael Spooner
Donal O’Dwyer
Ben-Zion Weiner
Chairman (Non-Exec)
MD & CEO
Non-Exec. Dir.
Non-Exec. Dir.
Non-Exec. Dir.
Major Shareholders
Silviu Itescu
Cephalon Inc
M&G Investment Funds
Thorney Holdings
24.0%
19.6%
11.3%
6.1%
Share Price Chart
Source: FactSet
Mesoblast (MSB) is an Australian Stock Exchange (ASX) listed
biotechnology company focussed on developing a suite of
products based on its proprietary adult stem cell technology. This
report relates to criticisms by another broker of Mesoblast,
predominantly regarding their clinical trials, and provides
opposing arguments to the points raised.
Background
A report recently circulated criticises Mesoblast for not publishing
results of clinical trials, engaging in post-hoc data analysis in relation to
a presentation to the American Heart Association (AHA) national
conference in 2011, omitting important data from that same
presentation and alleges that many trials have stalled.
Commentary
At this stage, there is no urgency for Mesoblast to publish its clinical
trial data, given they already have a partner for the cardiac and bone
marrow transplant applications of their stem cells. Their deal with
Cephalon (now Teva Pharmaceuticals) provides validation of the data
in those two areas beyond that which would be obtained by the peer
review process for journal articles, due to the more invasive nature of
the due diligence process in examining study results involved in
licencing transactions. Having said that, a search for the term
Mesoblast identifies 78 records in a popular database of scientific
articles and we expect this list to grow.
The AHA presentation was NOT a Mesoblast presentation, but rather a
presentation by one of the investigators involved in Mesoblast’s phase
II congestive heart failure trial and, consequently, it is difficult to
criticise Mesoblast for the handling of the data in relation to the
presentation. Regardless, the very nature of conference presentations
tends to ensure that they contain the appropriate information. It is also
standard practice for clinical trial data to be parsed in as many ways as
possible, because it is such a scarce and valuable resource. The
Cephalon deal, again, provides investors with assurance regarding the
quality of the data.
While one might conclude that progress in five indications has stalled,
four of the five indications are the subject of the licencing agreement
with Teva and news flow regarding these projects is under Teva’s
control. This is a standard part of just about any licensing agreement
between a small biopharmaceutical company and a larger company.
An early heart attack trial was abandoned and it is common knowledge
that the trial protocol made it difficult to recruit patients. The spinal
fusion trials are behind the original schedule, but Mesoblast has
indicated results should be available around the end of this year.
Further opportunities that have developed post-commencement of the
spinal fusion trials (e.g. diabetes and degenerative disc disease) are
attractive and would logically be a higher priority than spinal fusion.
Recommendation
Strong BUY rating and 12-month price target of $13.15 maintained.
Mesoblast
Lodge Partners Pty Ltd 2 Tuesday, 25 September 2012
Recently, a report issued by broker criticised aspects of Mesoblast’s clinical trials to provide
examples of negative signals investors should read into particular events or occurrences with
biopharmaceutical companies. While the report does highlight some pertinent points, the
examples it uses are simplistic and in many cases just wrong. The claims and an analysis of
them follow.
Firstly, the criticism was made that Mesoblast has published the results of none of its
human studies in the scientific literature.
There are principally two reasons why companies publish in scientific journals.
1) The first is simple, it is what scientists do. In a lot of ways, journal articles are to
scientists, what money is to a businessman. A biopharmaceutical company that did
not allow its scientists to publish would not be able to attract very good scientists.
2) Secondly, biopharma’s publish to provide marketing material. This material can be
used for a variety of reasons. Some of the common reasons are to provide the
company with credibility when seeking meetings with prospective partners, to support
product launch or to address perceived inadequacies of a product once it is on the
market.
These two interests will determine if and when an article is submitted for publication (a
company cannot control precisely when a study is published).
Given Mesoblast is well funded and has a significant partner in Teva Pharmaceuticals
(NYSE: TEVA) there is not a lot of reason for the company to publish key data in journals in
order to attract partners at this point in time.
By way of example, there appear to be only three papers in the published literature on
Acrux’s (ASX: ACR) and Eli Lilly’s (NYSE: LLY) product, Axiron, and all three have appeared
post the product’s approval by the US Food and Drug Administration.
An appropriate publication strategy is to build a body of research capable gaining publication
in a high impact journal, such as the New England Journal of Medicine, closer to the time that
it is needed. Articles in high impact journals provide extremely good marketing material.
There are many reasons not to publish, but they generally come back to one point:
Competitive Advantage. There is no reason to give your potential competitors anymore
information than you need to. Period.
For all the business theory, though, a search of PubMed using the term Mesoblast retrieves
78 records. PubMed is, to our knowledge, the largest freely available online database of
scientific literature in the world. Suffice to say, Mesoblast’s scientists have published a
significant amount of material and we expect they will publish further papers in high impact
journals.
Secondly, Mesoblast is criticised for engaging in post-hoc or retrospective data
analysis.
These terms refer to looking at scientific data after it is collected and using that data to guide
analysis of it. Post-hoc analysis falls under the broad heading of data mining. Data mining
refers to the practice of looking at scientific data in as many ways as conceivably possible,
whether the analyses were specified in the study protocol or not. Unscrupulous companies
will engage in this practice in an effort to make a negative experiment (e.g. clinical trial) look
positive when it was not.
Based on accepted scientific practice, one in every 20 analyses will return an erroneous
statistically significant result. That is, for example, a scientist will conclude a drug has had an
effect, when it actually hasn’t based on one in twenty analyses (on average). Do enough
analyses and you will find one result supporting the view that the drug works. Announce that
Mesoblast
Lodge Partners Pty Ltd 3 Tuesday, 25 September 2012
result alone and investors can be misled.
Data mining is a valid scientific method in the search for evidence to support a hypothesis
(e.g. that a drug prolongs life). BUT, in a strict scientific sense, if a finding is the result of data
mining, the experiment (or study) must be repeated with that particular variable as a primary
output (i.e. endpoint) before any conclusion can be reached. In the industry of biopharma,
decisions on whether to advance a product or not are, generally, made on a probability and
risk adjusted discounted cash flow type basis. A project advanced on the basis of a result
obtained by data mining attracts a lower probability of success than if the result was obtained
under strict scientific rigour and, consequently, attracts a lower net present value and has a
lower chance of being advanced by the company concerned.
Data mining typically occurs in clinical trials because trials are very expensive to run (trials
typically cost between $25 thousand and $100 thousand per patient, sometimes more) and a
company would be remiss if it didn’t try and extract as much knowledge about their product
out of a trial as possible.
In their note, the broker has criticised the handling of the data from Mesoblast’s phase II trial
of its stem cells in the treatment of congestive heart failure (CHF) in a 2011 presentation
made by Emerson Perin, MD, PhD at the American Heart Association’s national conference.
The first point to note is that this presentation is NOT Mesoblast’s. Emerson Perin is a
Principal Investigator and Director, Clinical Research for Cardiovascular Medicine; Medical
Director, Stem Cell Center at the Texas Heart Institute.
While companies are often involved to some extent in helping prepare presentations by
clinical investigators, it is the investigator who determines what is and what is not included in
a presentation. Moreover, at conferences, audiences are generally filled with experts, in this
case surgeons, cardiologists and scientists, with expertise in CHF, among other things. To
give a presentation in this environment where the data had been massaged, crucial pieces
left out or mined data given too much significance would instantly be noticed by the audience
and tarnish the reputation of the presenter. It is this potential for reputational damage that
acts as a strong driver to ensure speakers’ presentations are not unduly influenced by the
wishes of study sponsors (in this case, Mesoblast).
A second point to note is that Mesoblast has always been clear that the efficacy endpoints in
this trial were exploratory. As stated in the entry regarding the phase II CHF trial on
ClinicalTrials.gov, “The secondary objectives are to explore functional efficacy for subsequent
study design”.
For the reasons stated above, trials such as this one generally look at as many variables as
they can, in as many ways as is possible and these analyses are sometimes specified,
sometimes alluded to, in the trial protocol (not every communication regarding the trial).
There is nothing sinister about this. It is simply a company maximising the value of a scarce
resource.
If Mesoblast had claimed the results definitively demonstrate that their stem cells reduce
major adverse cardiac events in CHF patients, they could be accused of misleading
investors. They haven’t. They have only ever maintained that the results justify the study of
their stem cells in a later stage trial; a statement that appears perfectly justified by the data.
So, how solid is the data from the phase II CHF trial?
The good news for investors is that they do not need to undertake a specialist medical degree
and a PhD to arrive at an answer. Cephalon has largely done the hard work for them. The
process that a company like Cephalon goes through before they pay a $130 million up-front
fee, take a $96 million equity stake (the value at the time of the deal) and promise $1.7 billion
in milestones plus royalties is fairly exhaustive. While the peer-review process provides
validation for a company’s trial results, it is not nearly as comprehensive an examination of
the data as occurs in a deal such as this. Reviewers don’t go through the case report forms of
each patient enrolled in a trial or collate the results of a study independently of the company
before a study is accepted for publication.
Mesoblast
Lodge Partners Pty Ltd 4 Tuesday, 25 September 2012
Thirdly, Mesoblast is criticised for omitting data which doesn’t suit the story.
There is no question that some companies omit data which doesn’t fit or calls into question
results they have obtained with their products.
This criticism of Mesoblast, however, is based, again, on Dr Perin’s presentation at AHA in
2011. As stated before, the presenter has the final say as to what is and what isn’t included in
their presentations. It is difficult to take Mesoblast to task for that and, again, the nature of
conference presentations tends to ensure that they contain the appropriate data.
We have known there was a bias towards later stage CHF patients in the control group of the
phase II clinical trial based on some parameters for some time and have heard Mesoblast’s
CEO address the issue publically a number of times. It is worth noting that Cephalon would
have been aware of this bias before striking their deal with Mesoblast, suggesting they did not
see it as an issue. Small studies often have biases such as this and it is one reason why
larger trials are required before regulatory bodies will grant marketing approval for a product.
Fourthly, it is claimed that Teva’s upcoming decision on whether to fund a phase III
trial of Mesoblast’s stem cells will be the first rigorous test of Mesoblast’s phase II CHF
data.
This assertion is false. The first big test Mesoblast’s phase II CHF data faced was due
diligence by Cephalon in preparation for their deal with Mesoblast, the second was Dr Perin’s
presentation at AHA.
Also, it is incorrect to assert that clinical data is the only aspect a company looks at when
deciding whether to licence or fund development of a product. It is very important, but other
factors, such as the company’s therapeutic focus, are equally important.
Finally, it has been claimed that Mesoblast’s clinical trial progress has been somewhat
mixed.
Firstly, it is incorrect to suggest that fast progress in biopharma equates to good progress.
Competitive and treatment landscapes in biopharma change all the time and predicting these
changes is not an exact science. This can create situations, for example, where the relevance
of a clinical trial is compromised because the comparator being used in the trial no longer
represents the standard of care (SOC) in medical practice. In such situations, it is better to
cease the trial and move back to the appropriate point in product development to enable the
change in the SOC to be accounted for before re-commencing development. Continuing a
trial simply to support ‘appearances’ just isn’t logical.
Two of the trials mentioned in the other broker’s report are examples of rational product
development, not stalled progress.
The heart attack trial mentioned by the broker was suspended for some time before the
Cephalon deal was done. It has been common knowledge for a while that recruitment was
slow because the trial protocol stipulated the patients have a second procedure to receive
Mesoblast’s stem cells and few patients wanted to have another procedure after having
already been treated for the attack. Our understanding is that this issue has been rectified in
the current heart attack trial protocol.
While Mesoblast has been developing its spinal fusion project, two things have happened.
The first is that the SOC has changed. It has moved from a fairly invasive autograft (using a
patient’s own bone harvested from elsewhere in the body) procedure to a minimally invasive
procedure using bone morphogenetic protein (BMP) to stimulate bone growth and fusion.
Mesoblast’s original trials used the current SOC at the time, autograft. The new trials, which
have required Mesoblast to make additional regulatory submissions, incorporate the new
SOC. This will put Mesoblast’s product in a much better competitive position because their
stem cells are likely to have significant safety benefits over BMP.
Mesoblast
Lodge Partners Pty Ltd 5 Tuesday, 25 September 2012
The second thing that has happened is that additional potential applications of Mesoblast’s
stem cells have become apparent. Principally, these opportunities are diabetes and
degenerative disc disease. Given the size of these opportunities, spinal fusion has been
given a lower priority.
It is also important to note that four of the five product indications (bone marrow transplant,
CHF, heart attack and angina) the broker report uses to demonstrate that Mesoblast’s
progress has been mixed relate to indications licensed originally to Cephalon and acquired by
Teva in its acquisition of Cephalon. When a smaller company licenses a product to a large
pharmaceutical partner, the large pharmaceutical partner will almost invariably have complete
control over news flow regarding the licenced product. The point being, questions regarding
these indications aren’t Mesoblast’s to answer and large pharmaceutical companies tend to
give out as little information as possible. Investors invariably don’t like this, but it is routine
practice. Cheers Vin