Morningstar Magazine - August/September 2017 - 67

melanoma, head and neck, because they're
smaller. But when you add them up they
become very large. They represent in totality more
than the non-small cell lung cancer market.

that Roche's Tecentriq is well positioned-
probably not as well positioned as Bristol's or
Merck's immuno-oncology drugs, but nevertheless
reasonably well positioned.

I understand that these drugs are narrowly targeted,
so that a drug for lung cancer would not necessarily
be applied to head and neck.

This will help drive some of the growth that we're
anticipating for Roche. Roche is a pretty
diversified company. It's not a pure play on
immuno-oncology, which represents close to 10%
of total sales by 2021, by our estimate-
not an overly dominant piece of the overall sales,
but nevertheless an important contributor.

Conover: Cancer used to be just cancer, and one
chemotherapy would likely work for all, or at
least partially work for a lot of cancer. What we've
subsequently found is cancer can be broken
down by parts of the body and then can
be further broken down by biomarkers. Typically
speaking, when these drugs work for a type
of cancer with a certain type of biomarker, that
can sometimes translate into a benefit for
other cancers with a similar biomarker. However,
that's not always the case. There is some
variability of response rate across the different
types of cancer, and certain cancers can be
very different.

We're also going to see different types of trial
design that might enable different ways to
cut up the market even further. For example, with
non-small cell lung cancer, some studies
focus on certain biomarkers, some studies on
others. It's very likely we're going to see room for
a lot of different drugs within similar cancers
because of the way the trial design works, and
then that might lead physicians to use one
drug in one way and another drug in a different
way, for the same type of cancer.
To turn to specific investment opportunities,
Roche RHHBY is on your Best Ideas list. Why is this
name undervalued?

Conover: With Roche, there are a couple of key
dynamics going on. First off, they were one
of the first companies to launch biologic drugs,
and so now they're one of the first companies
to face potential biosimilar competition. There's
concern in the investment community that
the biosimilars will take a lot of market share from
Roche's older biologics, drugs like Herceptin
and Rituxan. We also think the biosimilars
are going to take some market share. But we think
people are underappreciating Roche's nextgeneration drugs, and one of the important pillars
there are immuno-oncology drugs. We believe

You also singled out Bristol-Myers Squibb BMY
in your report.

Conover: Bristol is more of a pure play on
immuno-oncology. We are bullish on immunooncology, so it makes sense that Bristol
might be undervalued. We expect almost half
of the sales by 2021 for Bristol will come from
immuno-oncology drugs.

One thing we like about Bristol relative to some
of the other players is that it has a first-mover
advantage into a lot of cancers. That helps
establish market share, helps create a moat within
the immuno-oncology space. Typically speaking,
the first mover to a particular therapeutic
area within cancer will dominate the market share.
A drug that's approved second or third to market
tends to get a very small slice of the overall
market share, even if the efficacy or the side
effects look a tad better. Doctors get comfortable
with the medicine and don't want to switch,
and you have longer clinical data, which helps
reinforce its positioning.
Bristol had a setback with the CheckMate-026 trial.
What were the lessons there?

Conover: That was a major setback. Bristol's
study did not statistically beat the standard of care
for first-line non-small cell lung cancer therapy.
We believe that's largely due to the design
makeup of the trial. Had this study been properly
stratified with the information that we know
now based on tumor mutational burden and other
factors, we think it's very likely the drug would
have worked. It was reaffirming to hear Dr. Seiwert
agree with the assessment that Opdivo-
the drug that failed in CheckMate-026-is a very
effective drug.

If you look at Opdivo's performance in almost all
the other studies, we're seeing positive
outcomes, especially in the second-line lung
cancer setting. That gives us confidence
that the drug will eventually work in the first-line
setting, in combination with another drug of theirs
that should enable them to get a slice of
that large non-small cell lung cancer market.
Do immuno-oncology drugs require different
approaches to trials?

Conover: Trial design is critical for drug development for immuno-oncology, and it's early
days here. There have been theories about
immuno-oncology for close to 100 years, but they
weren't put into effect because we haven't
been able to design the right type of drugs. About
eight years ago, we really saw the first emergence
of immuno-oncology drugs working, and
we've seen more since, but we're very early in
this process.

Immuno-oncology drugs don't work for everybody,
but they work for a very large portion of the
patient population. To correctly define that patient
population is still a bit hard because the
scientific community hasn't been able to find the
best biomarkers that will enable them to
run studies to gain the efficacy needed to show
statistical improvement. But we're much further
along now versus a little over two years ago,
when Bristol first set up that study that eventually
failed. Firms are taking a bit of a risk developing
phase 3 studies without a full information
set, because they want to be first to market. You
increase the risk of failure. But these drugs
are so strong that there is a little bit of a safety net
there, because the drugs typically work.
Merck MRK has a significant market share in
these drugs, but you noted it is not as good a value
at the moment.

Conover: Merck is probably going to get the lion's
share of the immuno-oncology market. They
already are first to market in first-line non-small
cell lung cancer, which again is the most important
market out there from size, and so that should
enable them to get a dominant share of the overall
immuno-oncology bucket. But the stock has
really run up to close to our fair value now. There
are some other things going on. By 2021, we

global.morningstar.com/Morningstarmagazine

http://global.morningstar.com/Morningstarmagazine

Table of Contents for the Digital Edition of Morningstar Magazine - August/September 2017