Holman_Nature biotechnology_gene patent article revised - AIPLA

ahemhootBiotechnology

Dec 5, 2012 (4 years and 6 months ago)

289 views

1



Debunking the Myth
That
Whole Genome Sequencing

Infringes Thousands of
Gene Patents

Christopher M. Holman J.D., Ph.D.

Associate Professor of Law

University of Missouri
-
Kansas City School of Law



Thanks to rapid advances in

DNA sequencing

technology, the cost of whole genome
sequencing
(“WGS”)
is plummeting, and it is anticipated that
the

much her
alded “$1000
Genome” will

soon

be

a reality.
Increasingly, however,

there is
concern
that
gene patents will
impede the implementation
of personal

genome sequen
cing
, to the detriment of multi
-
genic
diagnostic testing and personalized medicine
.

The fear
that human gene patents pose a
threat

to
WGS

is

based
largely on the following
widely
-
held misperceptions
:



The holder of a "gene patent" owns the pat
ented gene
, and any analysis of a patented
gene, including
DNA sequencing
,

constitutes
patent infringement
.



20% of

human genes are "patented,” and so sequencing an individual's whole genome
would

inevitably

result in the infringement of thousands of gene patents.



Ergo
, in order
to sequence an individual's

genome,

it will be necessary
to

obtain licenses
under thousands of gene patents, which will be prohibitively expensive (some have
suggested $100,000 per ge
nome), or

risk numerous costly pat
ent infringement lawsuits

(1)
.

While the foregoing chain of inferences might

appear plausible on its face, it is in fact based
on
several

seriously

flawed assumptions
.

This article

explains

why
the actual

threat

posed by
gene patents

to
WGS

has likely been greatly overstated.

No one can
"own"

a gene

The term "gene patent"
is

something
of a misnomer, and it is important to recognize that
gene patent holders do not

actually

“own”
the
genes

which are the subject of their patents
.

The
only right

conferred by a patent is the right to exclude others from making, using, selling or
importing
a
non
-
naturally occurring, man
-
made

product or process, as defined by the patent
’s

"claims."






This research was conducted
with financial support from the Biotechnology Industry Organization (BIO). The
design, analysis and composition of this manuscript was conducted independently by me, and I am responsible for
any errors or omissions. The views and conclusions expressed here
in are my own, and do not necessarily reflect
the views of BIO or any of its members
.

2


The claims
appear the end of the patent and formally define the scope and extent of the
patented invention.
When it comes to patents
, "the name of the game is the claim”
-

it is
impossible make any sort of valid assessment of the

scope
or impact
of a patent without r
eading
and interpreting the
claims
. Unfortunately, much of the
commentary on
gene patents

has failed to
focus on

the

language of the claims,
leading

to unwarranted assumptions as to the impact of gene
patents on
genetic diagnostic testing, and in particula
r on genome sequencing
.

To begin with
, the law is
quite

clear that genes, as they occur

naturally, e.g., in the human
body, cannot be
patented
.
Neither can

genetic information, such as the DNA sequence of a gene,

or a correlation between a genetic
mutation or
variation and
disease
. It is
, however,
possible

to
claim

non
-
naturally occurring, man
-
made DNA

molecule
s
. Thus, for example, the Court of
Appeals of the Federal

Circuit
(“Federal Circuit”)
recently affirmed
in
Association for Molecular
Patholog
y v. US Patent and Trademark Office
(“
AMP v. PTO
”)
that an "isolated" DNA molecule
that corresponds in sequence to a naturally occurring gene can be patented

(1)
. This ruling
is
consistent
with
the

longstanding

position of the

US Patent and Trademark Office

(“PTO”)
that

isolation of
a biomolecule from

a natural source can constitute sufficient human intervention to
render the isolated molecule patent eligible

(2)
.
Patent protection is also available for
inventive
methods

of
us
ing isolated
DNA molecules
,
for example in recombinant protein expression, and
for
methods of analyzing genes and genetic variation, such as in a

genetic diagnostic test.
“Gene
patent
" is

simply

a

convenient

label for patents

that include
claims of this
type
, but its usage has
unfortunatel
y

led many to believe incorrectly that genes
per se

can be patented
.

The only

gene patents that c
ould be infringed by
WGS

are those that include
a claim that
encompasses a
non
-
naturally occurring
product
(i.e., a
n isolat
ed or synthetic

DNA molecule)
or
a
process that would be

inherently

made or used in the cou
rse of
WGS. It is incorrect
to assume
that all
“gene
patents


would
necessarily
be infringed by
WGS without
interpreting

the claims
and comparing the
properly interpreted
claims to the specific WGS methodology contemplated
.

For
example
,

C
laim

20 of US

Patent No. 5,747,282, one

of
Myriad Genetics’ patent
claims
recently

chal
lenged by the ACLU in

AMP v. PTO
,

only covers

methods of using the
"patented" BRCA gene in cell
-
based assays to identify
potential cancer therapeutics

(3)
.
Although the ACLU asserts that it brought the lawsuit in order to improve access to BRCA
genetic testing, this particular claim
clearly
could not

be infringed by
DNA sequencing
, or
for
that matter
by any form of genetic diagnostic testing.
In fact,

as described below,

my analysis of
a large number of gene patent claims strongly suggests
that
most

gene patents

are unlikely to
include a single claim that would
be infringed by
WGS
.

The
20%
myt
h

The
statistic

that 20% of human genes
are patented

has become something of an urban
legend, and is routinely cited as
though it were
an
established
fact, often w
ithout any supporting
reference.

A

thorough

review of the literature
, however,

reveals that

this misperception

stems
3


from a single

“Policy Forum”

article
published in Science
in 2005
by Jensen and Murray
purporting to map

the "intellectual property landscape

of the human
genome
” (
“the
Jensen/Murray

a
rticle

“)
(4).

A

careful reading of
the article

reveals
that it

has been widely
misinterpreted, and it
in

no way supports
a conclusion

that

20% of human genes are patented in a
manner that would be infringed by
WGS

or genetic diagnostic testing.


The main text of
the Jensen/Murray
a
rticle

does indeed state
that "20% of human genes
are
explicitly claimed

as US IP,"

and this has no doubt
led many to infer that
20% of human
genes are exclusively
"
owned
"

by patent holde
rs. However,

the

Supporting Online Material

for
the article
clarifies

that

what

the authors actually found was

that
,

with respect to 20% of
the
human genes
which had been identified

at that time, at least some portion of the
gene’s
DNA
sequence, or the amino acid sequence of the corresponding protein,
was
explicitly mentioned

in
a
US patent claim

(5).

The myth that 20% of human genes are pa
tented has taken root because so
many have failed to appreciate the
critical distinction between
a DNA or
amino acid sequence

being "mentioned" in a patent claim and a gene being "claimed.”
The mere fact that a patent
claim mentions a gene sequence,
or the

sequence of the

protein encoded by it,

in no way implies
that the
patent completely precludes access to the gene, nor does it warrant an assumption that
sequencing the gene would result in patent infringement.

The Jensen/Murray

claims

In order to

meaningfully

assess

the significance of
t
he

Jensen/Murray
article
,
I
took
it
upon myself

to read the claims
of

the patents which form the basis for the article’s
assertion

that
"20% of human genes are
explicitly claimed

as US IP."
To
this end
, I asked
one of the article’s
authors

(Jensen)
to provide me with a listing of the 4270

"gene patents" they had identified in
their study,
and

he

graciously
complied

with my request
(6).

T
he identity of these patents
had
never pre
viously been published, and
Jensen

informed

me that
in fact
I was the first person

to
ever ask

to see the list.
Each patent

contains multiple patent claims, sometimes hundr
eds of
claims
, so to make my task manageable I randomly selected 533 of the 4270 patents

and
reviewed

all of the claims in each of these patents

(7)
.

At this point it bears
emphasizing

that

the interpretation of
patent claims
, particularly

outside the context of
patent
infringement litigation
,

is a notoriously unpredictable
undertaking
.
The vagaries of claim interpretation have been well documented, and are reflected in the high
rate at whic
h the

Federal Circuit reverses the claim interpretation rulings of
lower

court
s

on
appeal
. It

is not uncommon for the Federal Circuit judges to disagree amongst themselves as to
the p
roper interpretation of a claim
.

Indeed, the ambiguity of the scope of c
overage defined by
patent has led some, including at times the Federal Trade Commission, to characterize patents as

mere
"probabilistic" property rights.

The uncertainty is especially pronounced with respect to gene patents, which have rarely
been litigate
d, particularly in the context of genetic testing,
and
as a consequence of which there
4


is little guidance to be found

in the case law (8
).

Nonetheless
,

it is possible

for
a patent attorney

such as myself,
familiar with the legal doctrines of patent claim
interpretation and the science of
mo
lecular biology,
to
read and construe the claims of these

patents
and
at least
a
rrive

at

an
informed assessment as to the likelihood they would be found infringed by
genetic testing, and
more particularly by
WGS
.

I found that
more than one
-
quart
er

(
139
/
533
) of the

sampled patents

do not claim a DNA
molecule corresponding in sequence to naturally occurring gene, nor do they claim any method
of analyzing
a
gene

(7)
.

This subset of the

patents

probably should not even

be considered gene
patents. For example,
most of these patents are

directed solely towards proteins,
not genes or
DNA,
and their identification as gene patents in
t
he
Jensen/Murray

article

arises as an artifact
of

the automated query
those authors

used to identify gene patents

(9)
.
Other patents

in this subset

include claims that
reference

DNA sequences, but
only cover

inventions such as fusion genes,
vaccines, or methods of DNA processing and analysis

that are

totally unrelated to sequencing or
ge
netic analysis.

One striking example is
US patent number 6,500,938,
in which the

broadest claim covers
“[a]

combination comprising a plurality of polynucleotide probes, wherein said plurality of
probes are SEQ ID NOs:1
-
1490.” This claim does not recite any

full
-
length gene sequence, only
specific probes derived
from portions of cDNA sequences
.

The claim is clearly directed towards
microarrays,
and would only be infringed by a microarray (or
some functional

equivalent of a
microarray)
that includes

each and every of these 1490 specific probes sequences.
Such a patent
has

absolutely no implications for
WGS
, or for the vast majority of research or diagnostics uses
of
the
genes
represented by

the probe sequences
. Nonetheless
, this single microarray cl
aim, by
virtue of the fact that it mentions probes representing up to 1490 human genes, could account for
up to one third of the human genes identified by Jensen and Murray as "mentioned" in patent
claims
,

and which
thus
form the basis for the myth that 20
% of human genes are
“patented.”


WGS

unlikely

to infringe

The remaining 394

patents in the sample

I analyzed

include claims

that more
plausibly
qualify as gene patent

claim
s,

and

with respect to
these patents

it is more difficult to
categorically

rule out the possibility of infringement by at least some forms of
WGS
.
These
patents include claims falling into
two broad categories: (1)

composition
-
of
-
matter
(“COM”)
claims

directed
towards
DNA molecules that correspond in sequence to at least some po
rtion of
a

human gene, and (2) method

claims relating to

the analysis or detection of

specific

gene
sequences or
genetic variations. Nonetheless, in spite of
some degree of uncertainty as to exactly
how a court would
interpret these claims, my analysis leads me to conclude that few if any of
these patents include claims that would likely be
found infringed by
WGS
, particularly next
-
generation

methodologies that do not rely on DNA amplification

(10)
.

5


369

of the
patents

include a claim that falls into the first category
, i.e.,
COM

claims.
See,
for example, Claims 1 and 2 of US Patent No.

5,342,744

(7)
.
These

claims
would only be
infringed by a DNA sequencing methodology that necessarily entails "making or using" the
DNA

molecule

as defined by the claim. There is a great deal of heterogeneity with respect to the
actual language used in these claims, so it is difficult to make
sweeping generalizations
,
but for a
variety of

reasons it appears that
, as a general matter,

the
se claims will prove little obstacle to
WGS
.

To begin with,
claims to DNA molecules having the sequence of a naturally occurring
human gene are almost invariably limited to forms of the DNA that are
"
isolated
” (sometimes
synonymous limitations such as "pur
ified" are used)
-

without such a limitation
,

the claim would
be invalid for encompassing the gene as it occurs in nature.
To my knowledge, after extensive
research, no
U.S.
court has ever interpreted a claim to an isolated or purified DNA molecule so
broa
dly

that it would be

infringed by DNA sequencing. In fact, there

is judicial precedent that,
while not directly on point,
would support a relatively narrow interp
retation of isolated DNA
claims, such that these claims
would not be infringed by at least som
e forms of DNA sequencing,
particularly certain next
-
generation technologies that do not r
equire the amplification of DNA

(11)
.

For example, the Pacific Biosciences
SMRT™
gene sequencing technology relies on the
observation of DNA synthesis as it occurs on an immobilized DNA polymerase,
and thus
arguably does not entail

iso
lat
ion of defined DNA molecules (12
).

Some commentators have voiced concern that a court might interpr
et the term "isolated"
very broadly
, so as

to encompass DNA occurring in any context distinct from its native
environment

in the genome of a

naturally occurring human cell. For example, Robert Cook
-
Deegan

of Duke University has opined

that
observation
of D
NA synthes
is on a tethered DNA
polymerase inherently

entails isolation of DNA molecules
,

because the polymerase is hydrogen
bonded to a particular DNA molecule during the course of sequencing

(13)
.
Moreover, he

believes that any technology for reading a DN
A sequence will necessarily involve isolation, and
thus

constitute

infringement if the term "isolated" is interpret
ed broadly, a possibility he feels
cannot be ruled out
.

Because no court in the US

has ever addressed
this iss
ue head
-
on, it is impossible
to
entirely
rule out the possibility that a court

would
interpret a claim to an isolated DNA molecule

in the extremely broad sense suggested by
Cook
-
Deegan
.

However,
were
it
to do so, I think it is
highly likely that the court would also find the claim inv
alid. Accused infringers typically
challenge the validity of patent claims that have been asserted against them, and it is routine for
the courts to invalidate claims during the course of patent litigation. As a general matter
, the
broader a claim is inter
preted the more likely it is to be invalidated for failure to comply with one
or more r
equirements of patentability.

For example, in order to be patentable a claimed invention must be novel and
nonobvious. The actual legal analysis would be quite complex,
but the fundamental principle

at
6


stake is quite a
straightforward: subject matter that would infringe a patent claim after the patent
issues will invalidate the patent if

that subject matter was publicly used, or described in a printed
publication, more th
an one year before the filing

date of the patent

(14
)
.

Thus, if an isolated
DNA claim is interpreted so broadly as to cover a DNA molecule

non
-
covalently
bonded to
a
tethered DNA polymerase,

for example,
the claim could presumably be invalidate
d by
showing
that prior to the patent

a

DNA molecule

covered by the claim

had been non
-
covalently bonded to
a solid surface. But this sort of

isolation


of DNA has
been described

for many years, e.g.,
using techniques such as Southern blotting

(which involves

the separation and immobilization of
fragments of genomic DNA)
,
in publications
pre
-
dating all currently active gene patents
,
suggesting that under such a broad interpretation the claim would be found invalid
.

Another

example

of "isolated" DNA,
w
ere the
term to be

interpreted broadly, would
be
genomic DNA libraries, which like Southern blotting have been used extensively in molecular
biology since the 1970s
, and inherently involves

the isolation of la
rge fragments of genomic
DNA
.

The announcement of

a working

draft
of the

human genome in 2001, and the completed
sequence in 2003,

is particularly significant in the context of whole genome sequencing
, because
it strongly suggests that any patent filed
since then
cannot be both valid and infringed by WGS
,
at least by the conventional methodologies use
d in
the initial sequencing of the

human
genome.

Cook
-
Deegan agrees that, if interpreted very broadly,

isolated DNA claims

are probably
invalid

(13)
. However, he believes that because the courts have not yet
had the opportunity to
address the validity of these claims, the outcome is uncertain, and this uncertainty is
itself
problematic. This might be true, but uncertainty as to patent scope and validity prior to litigation
is not limited to gene patents, and I

see little reason to think that it will prove unduly problematic
in the context of
WGS
.

It bears noting that the

reason we have so little guidance from the courts as to the scope
and validity of gene patents in the context of genetic testing is that
these

patents have
rarely been
asserted against the providers of diagnostic tests

(8)
. To my knowledge, there
has only been one

case in which
the enforcement of a gene patent in the context of genetic diagnostic testing has
resulted
in
a judicial decision, and

in that decision all of the asserted gene patent claims were
declared invalid,
an outcome

consistent with
the thesis of this article

(15
)
.

Furthermore, even if

a court

were to interpret
the term "isolation" very broadly, such that

a mode of DNA sequencing

would

necessarily entail the “isolation” of a claimed DNA molecule,
there are other limitations

inherent in most COM
claims that would appear to preclude
infringement by

at least some formats of

DNA sequencing. For example,

most
of the patents

in
the sampling

I analyzed were based on the isolation of cDNA, and
only include claims covering
the isolated cDNA molecule
.
If a patent owner
succeeded in convincing a

court
that
its

patent
claim coves
d genomic DNA
,

b
ut the patent

specificationn

only di
scloses cDNA
,

the

court would
likely

declare
the
claim

invalid

based on inadequate disclosure
, since most human genomic DNA
includes i
ntrons not present in the cDNA

(14).

7


Significantly, any DNA
molecule
"isolated" in the course of genomic sequencing would

be genomic DNA, including introns, and thus
generally
not

covered by a claim limited to cDNA.
The US

government apparently shares this view
-

during arguments before the Federal Circuit in
AMP v. PTO

an

attorney representing the
government stated that a
"vast majority" of the claims
to isolated DNA that have been issued by the PTO are to "cDNA, recombinant DNA, process
claims and the like," and hence would not cover genomic DNA (
nor
, by
imp
lication, genome
sequencing) (16
).

Furthermore, most of the isolat
ed DNA sequences claimed in the patents I analyzed
are

too long to be infringed by many forms of
WGS
.

Large
-
scale DNA sequencing protocols (of the
type that would be used for
WGS
)

typically
involve relatively short sequencing reads, which

at
most
necessit
ate the isolation of relatively short fragments of genomic DNA

(10)
. In contrast,
the vast majority of the patents in the sample
appear to
only include claims to full
-
length protein
encoding sequences, which would not cover the
shorter
DNA fragments
generated

in many (if
not all)
WGS

methodologies

(7)
. Some of the patents include claims to DNA molecules
encoding less than the full
-
length protein, but typically these fragment claims only cover DNA
fragments long enough to encode a polypeptide retainin
g the function of the full
-
length protein,
which would still typically be longer than the DNA molecules analyzed in
WGS

(7)
.

A relatively small subset of the patents

I
analyzed

include claims to smaller fragments of
isolated

(7)
.

Many of these patents specify that the claim fragments are probes or PCR primers,
which would most likely pr
eclude infringement by
most

modes of whole genome sequencing.

Some of these claims recite any fragment

capable of encoding a peptide

large enough

to elicit an
antigenic response, while others recite
even smaller fragments, e.g.,
a few of the patents purport

to cover any isolated DNA molecule comprising
a short stretch of
contiguous bases appearing in
the claimed gene sequence

(see, e.g., Claim 1 of

US Patent Number 5,559,023).

One might argue that some of these fragment claims might be inadvertently infringed by
the creation of DNA fragments during the course of at least

some modes of DNA sequencing.
However, the sheer breadth of
these claims, if i
nterpreted so as to cover genomic sequencing,
would very likely be their downfall
-

a recent article suggests that fragment claims of this type
could be declared invalid based on anticipation by a large number of prior art DNA sequences
fallin
g w
ithin the
scope of the claim (17
).

The expansive breadth of the claim would also raise
substantial issues of patentability under the enablement and written description doctrines,

essentially

for claiming much more than
was

actually invented

(1
4
)
.

47

of the patents
I analyzed included one or more method claims directed towards some
aspect of genetic testing

(7)
. As with the DNA product claims, there is a great deal of
heterogeneity in the language of these claims, and owing to the inherent uncertainty involved in
interpreting patent claims it is impossible to categorically rule out the possibility that at least
some of these patents might be found infringed by
WGS
. However, there is good reason to think
that few
,

if any
,

of these patents would necessarily be infringed by
WGS
.

8


Many of these patents only

include claims directed towards
methods for determining the
p
resence o
f a specific genetic variation
by detecting hybridization to a nucleic acid or probe
.
See, e.g., US Patent Nos. 5,580,727 and 5,756,288
. It seems likely a court would interpret many
(if not all) of these claims as limited to methods of genetic te
sting that
involve direct detection of
a
hybridization event, and thus not encompassing DNA sequencing methodologies that only
incidentally

involve DNA hybridization, for example
, in the course of PCR amplification
.

Other
claims in this category

of patent
s

specifically require a PCR amplification step in order for there
to be infringement, which would
seem to rule out infringement by

DNA sequencing technologies
that do not require PCR amplification, such as some next
-
generation technologies capable of
s
equenci
ng a single DNA molecule (11, 1
8
).

Twelve of the patents include claims
that purport to cover detection of a
specific

genetic
variation by any means, without any explicit methodological limitations

(7)
.

See, for example,
US Patent Nos.
5,830,649

an
d 5,916,748.

Of all the claims analyzed in this study, these are
probably the most likely to be infringed by many

(
and
perhaps even all)

forms of
WGS
.
However, as I interpret these claims, they would most likely only be infringed by a singl
e entity
that
not only performs

the
physical sequencing of DNA molecules
, but also

analyzes the
resulting sequence data and recognizes
the existence of a particular polymorphism or genetic
variation

associated with disease,

as
specified in the patent claim.

For example
,

in a scenario where
a first entity

(e.g., a genome sequencing company)
sequences the whole genome of a patient,

thereby generating raw sequence data
, but does
not
analyze

the data

for the presence of a specifically claimed genetic variation, and the raw d
ata is
provided to
a second entity (e.g., a healthcare provider)

who performs analysis,

neither entity
would have performed both of the steps
required to infringe

the claim, and
in all likelihood
no
one would be liable for patent infringement
.

This illust
rates a poi
nt made by Myriad Genetics in
two
"friend of the court" brief
s

filed
recently with

the Federal Circuit
-

under current US patent
law, it appears likely that

there is no effective way to patent the discovery of a genetic variation
useful in diagn
ostics or personalized medicine
in a manner that would cover

a situation where
the sequencing of the DNA molecules and the analysis of the resulting genetic information are
performed by two independent entities

(
19
)
. In fact,
as technology advances,
the
more pressing
policy concern with respect to patents and genetic diagnostics might be that there is insufficient
patent protection available for such inventions, rather than too much patent protection.

Most gene patents
were
never

intended

to cover genetic

testing

Much of the concern over gene patents and genetic testing has been based on studies
,
such as the
influential
SACGHS Report

(2
0
)
,

that

have

focused on

the small subset of
gene
patents associated
a few

Mendelian diseases that are currently the subject of established genetic
diagnostic t
esting protocols. I
n 2009
, for example,

Huys et al.

reported that a number of gene
patents associated with

22 common genetic diagnostic tests
include claims that would be

"almost
impossible to circumvent"
by one performing

"currently prac
tic
ed genetic diagnostic tests” (2
1
).
9


Based on the

"almost impossible to circumvent”
language used in the article, some have

inferred
that these patents are virtually impossible to design
around, and would
be infringed by

any
format of genetic diagnostic testing, including
WGS

(2
0
, 2
2
)
.

However, this is an unwarranted assumption. As I confirmed by speaking with
Huy
s (the
lead auth
or

of the article
)
,
she and her co
-
authors

only compared the

patent claims

they analyzed

to the best practices guidelines for established genetic testing protocols. They did not consider,
or come to any conclusion with respect to, the likelihood the claims would be infringed by
alternate formats of genetic diagnos
tic testing, or by next
-
generation genomic sequencing
technologies. In fact, I conducted my own independent analysis of some of the claims flagged
by
Huys et al.
as "almost impossible to circumvent,” and found
many of
them readily susceptible to
circumvent
ion using alternate testing protocols.

Furthermore, it is
important to recognize that the

subset of gene patents relating to known
Mendelian conditions is not at all representative of gene patents in general. Patents
on

Mendelian
genes, not surprisingly, i
nclude patent claims directed towards genetic diagnostic testing,
including method cla
ims to genetic tests, and often
times claims to

short DNA fragments and
probes
likely to be employed

in some genetic testing protocols

(20, 21)
.

In contrast
,

the vast
maj
ority of the
Jensen/Murray
gene patents I looked at were
clearly
not drafted with genetic
testing in mind, and
as a consequence
do not include
claims

likely to encompass genetic
diagnostic testing

(7)
.

Rather, the majority of these patents arose out

of
the discovery of a
protein
-
encoding cDNA, and the perceived value of the
invention was the

use

of the cDNA

in
the production of recombinant protein, not as the subject of genetic testing.

The Incyte patents are a good example of this.
Incyte

Pharmaceutical
s is the assignee of
more gene patents in the
Jensen/Murray data
set than any other single entity

(
4,
6)
, and the
set of
claims in most of its patents follow a similar pattern: a claim to an isolated cDNA molecule
encoding a full
-
length protein, a claim to

an expression vector comprising the cDNA, a claim to
a host cell comprising the expression vector, and a

claim to a

method of using the cultured host
cell to express the full
-
length protein

(e.g., US Patent No.
5,994,112)
. Clearly, the intent of
this
clai
m set is

to cover use of the cDNA to produce a recombinant protein, not sequencing of the
gene or genetic diagnostic testing.

Gene patents of this type
, with claims
focused on use of the gene in

recombinant protein
expression,
are only as

sifgnificant

as the
encoded
protein. If the protein is useful as
pharmaceutical, then the gene patent can be quite valuable, as exemplified by Amgen and its
gene patent claiming the human erythropoietin gene

(23)
.

For most genes, however,
a gene

patent has little valu
e.
It seems apparent that the strategy of
Incyte
, and many gene patent
holders, was to obtain patent protection covering the recombinant expression of newly
discovered genes, in the hope that they would hit the jackpot and end up owning a patent
covering t
he production of a blockbuster biologic drug such as Epogen.

10


As of yet
, the gamble does not appear to
have paid off.

As of
April 11, 2011
,

only 37 of
the 398 Incyte patents flagged as gene patents in
the Jensen/Murray

article

were
still in force
-

the ot
hers

had
all expired owing to Incyte’s failure to pay the necessary maintenance fees

(
2
4
).
The early expiration of these patents is quite significant
-

Jensen and Murray reported that
Incyte’s “IP rights cover 2000 human genes,” which would represent close to half of the genes
identified in their study as
"explicitly claimed as US IP
"

(4).

In November 2011
,

Incyte received
its first FDA approval to bri
ng a drug to market


ironically, it is a small molecule drug, and thus
not

protected by any o
f the company’s gene patents (2
5
).

Conclusions

The assumption that 20% of human genes are patented
, and
that
as a consequence
WGS

will inevitably result in the in
fringement of thousands of gene patents, is based on a gross
misinterpretation of a single
“Policy Perspective” article
.


My analysis of the claims from
a
substantial sampling

of the

patents
which form the basis for the Jensen/Murray article indicates

that the vast majority of these patents

would almost certainly not be

infringed by
WGS
, either
because they are not gene patents at all, or because they only claim isolated DNA molecules
unlikely to be produced in
WGS

(e.g., long sequences and/or cDNA seq
uences), or methods of
genetic testing that would not encompass
WGS
.

There is also good reason to think that even the claims most likely to be infringed,
reciting short fragments of genomic DNA, or broadly defined methods of testing for genetic
variation,

would not necessarily be infringed by all forms of
WGS
, particularly next
-
generation
technologies that do not require gene amplification. A company

that provides
WGS

services, but
that leaves

the job of analyzing the sequence data for clinically significan
t variations to others,
would be particularly unlikely to be found liable for infringing any of these gene patents.

Furthermore,

there would seem to be
little incentive for the owner of one of

a gene patent
to sue someone for performing
WGS
.

At most, these

patents cover only a minute fraction of the
entire genome, so

even in the unlikely event a court were to find infringement,

it is highly
unlikely that
an

injunction
would be entered
barring the infringer from performing
WGS

(2
6
)
.
Likely the only remedy
that w
ould be available for the patent holder would be the award of a
"reasonable royalty" for use of the patented technology, which would
most likely

be
miniscule
considering the
relative
is
contribution of the patented invention to the sequencing of the
entire
genome

(2
6
)
.

In short, there appears to be little evidence that would support an assumption that gene
patents
pose a

substantial impediment to
WGS
. Ironically, it might be the case that insufficient
patent protection for genetic diagnostic discoveries exists to adequately incentivize their
commercialization, particularly if the Food and Drug Administration (FDA) takes on a more
active role in regul
ating genetic diagnostic testing, and begins requiring submission of data
demonstrating safety and efficacy, or if the Centers for Medicare and Medicaid Services or
11


private health plans and insurers begin to demand clinical studies before the
y cover and
re
imburse

for genetic tests

(18
)
. Without adequate patent protection, it might be difficult to
justify the substantial investment necessary

to bring the next generation of genetic diagnostic
products to market.

References

1.

Association for Molecular Pathology

v. U.S. Patent and Tradem
ark Office
, 653 F.3d
1329

(Fed. Cir. 2011)
, available at
http://www.patentdocs.org/2011/07/federal
-
circuit
-
issues
-
decision
-
in
-
am
p
-
v
-
uspto.html

(last visited
January 16, 2012
).

2.

Utility Examination Guidelines, 66 Fed. Reg. 1092 (January 5, 2001).

3.

US patents are freely available online from a number of sources, such as the US Patent
and Trademark Office website (
http://patft.uspto.gov/
) and Google Patent
(
http://www.google.com/patents
).

4.

Kyle Jensen & Fiona Murray,
Intellectual Property Landscape of the Human Genome
,
Science
310
:
239
-
40

(2005).

5.

More specifically, they conducted an automated search to identify all US patents reciting
the canonical term "SEQ ID NO.” in the claims, and wherein the “SEQ ID NO.” term is
used in conjunction with a specific genetic sequence corresponding to
a known human
gene
. Jensen & Murr
ay’s Supporting Online Material
, available at
http://www.sciencemag.org/cgi/data/310/5746/239/DC1/1


(last visited
January 16,
2012
).

6.

The entire dataset of patents which form the basis for the Jensen and Murray article
is
available

here (posting authorized by Kyle Jensen):
ht
tps://docs.google.com/spreadsheet/ccc?key=0At_llJGo9WK0dG00dUh4WXlNcm5UeF
pKcTNtMEhraVE&hl=en_US

.

7.

An annotated list of all 533 patents analyzed, including full claim sets, is available here:

https://docs.google.com/open?id=0B9_llJGo9WK0NjdmMmU3ZDctYzVkYS00OWQ5L
WI0NTktM2IwNTUxMDgzZDNl

.

8.

Christopher M. Holman
,
Trend in Human Gene Patent Litigation
,
S
CIENCE

322:198
-
99
(2008).

9.

The query employed by Jensen and Murray identified as "gene patents" patents that
claimed proteins by reference to the DNA sequence encoding the protein, or by reference
to a SEQ ID NO. that identified
not only

a DNA sequence
, but also
the amino acid
seque
nce encoded
by the DNA sequence
.

10.

Another commentator recently arrived at essentially the same conclusion. See
W.
Nicholson Price II,
Unblocked Future: Why Gene Patents Won’t Impede Whole
-
Genome
Sequencing and Personalized Medicine
, Cardozo Law Review Vol.
33, No. 4 (2012).

11.

Christopher M. Holman,

Will Gene Patents Impede Whole Genome Sequencing?:
Deconstructing the Myth That 20% of the Human Genome Is Patented
, 2

IP

T
HEORY

1
(2011).

12


12.

Single Molecule Real Time (
SMRT
TM
) DNA Sequencing, Pacific
Biosciences Website,
http://www.pacificbiosciences.com/assets/files/pacbio_technology_backgrounder.pdf
(last visited
January 16, 2012
).

13.

Personal communication.

14.

Brief of
Amici Curiae

Christopher M. Holman and Robert Cook
-
Deegan,
Association for
Molecular Pa
thology v. US PTO
, Federal Circuit Docket Number 2010
-
1406, available at
https://docs.google.com/viewer?a=v&pid=explorer&chrome=true&srcid=0B9_llJGo9W
K0OWVkYjg3NjItYzRjYi00ODIyLWIyMjAtYmJkZDQxMGZmYTZi&hl=en_US

(last
visited
January 16, 2012
).

15.

Billups
-
Rothenberg, Inc. v. Associated Regional and University Pathologists, Inc.,

642

F.3d 1031 (Fed. Cir 2011)
, available at
http://www.patentlyo.com/files/10
-
1401.pdf

(last
visited
January 16, 2012
).

16.

Oral argument in
Association

for Molecular Pathology v. US Patent and Trademark
Office
,Federal Circuit Docket Number 10
-
1406 , available at
http://www.cafc.uscourts.gov/oral
-
argument
-
recordings/search/audio.html

, at 55
-
56
minutes into the proceedi
ngs

(last visited
January 16, 2012
)
.

17.

Thomas B. Kepler, Colin Crossman and Robert Cook
-
Deegan,
Metastasizing patent
claims on BRCA1
, Genomics (2010), doi:10.1016/j.ygeno.2010.03.003.

18.

Kevin Davies,
The $1,000 Genome: The Revolution in DNA Sequencing and the
New Era
of Personalized Medicine

(2010).

19.

Holman’s Biotech IP Blog,
Myriad Genetics Files Amicus Briefs in Joint Infringement
Cases Akamai and McKesson

(July 24, 2011), available at
http://holmansbiotechipblog.blogspot.com/2011/07/myriad
-
genetics
-
files
-
amicus
-
briefs
-
in.html
.

20.

Report

of the Secretary’s Advisory Committee on Genetic
s, Health, and Society

on Gene
Patents and Licensing Practices and Their Impact on Patient Acc
ess to Genetic Tests
(April
, 2010) available at
http://oba.od.nih.gov/oba/sacghs/reports/SACGHS_patents_report_2010.pdfpdf (last
visited January 16, 2012).

21.

Isabel
le Huys et al.
Legal Uncertainty in the Area of Genetic Diagnostic Testing
,
Nature
Biotechnology
27:
903 (2009).

22.

Robert Cook
-
Deegan and Christopher Heaney,
Patents in

Genomics and Human
Genetics
,

Annual Review of Genomics and Human Genetics
11:
383 (2010).

23.

Amgen v. Chugai
, 927 F.2d 1200 (Fed. Cir. 1991).

24.

http://portal.uspto.gov/external/portal/pair (last visited January 16, 2012).

25.

Nuala Moran,
Incyte Comes of Age with JAK Inhibitor A
pproval
,

Nature Biotechnology
30:
3
-
5 (2012).

26.

Christopher M. Holman,
Gene
Patents under Fire: Weighing the Costs and Benefit
s,
book chapter in
B
IOTECHNOLOGY AND
S
OFTWARE
P
ATENT
L
AW
:

A

C
OMPARATIVE
R
EVIEW ON
N
EW
D
EVELOPMENTS
,
edited by Emanuela Arezzo and Gustavo Ghidini,
13


Edward Elgar Publishing (2011).
Draft manuscript
a
vailable
at
http://ssrn.com/abstract=1710150

(last visited January 16, 2012).