Stefan Bassarab et al.

Patent Trials and Appeals Board

Sep 25, 2020

13494398 - (D) (P.T.A.B. Sep. 25, 2020)

UNITED STATES PATENT AND TRADEMARK OFFICE
UNITED STATES DEPARTMENT OF COMMERCE
United States Patent and Trademark Office
Address: COMMISSIONER FOR PATENTS
P.O. Box 1450
Alexandria, Virginia 22313-1450
www.uspto.gov
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO.
13/494,398 06/12/2012 Stefan Bassarab A0871.70004US01 7431
23628 7590 09/25/2020
WOLF GREENFIELD & SACKS, P.C.
600 ATLANTIC AVENUE
BOSTON, MA 02210-2206
EXAMINER
GAMBEL, PHILLIP
ART UNIT PAPER NUMBER
1644
NOTIFICATION DATE DELIVERY MODE
09/25/2020 ELECTRONIC
Please find below and/or attached an Office communication concerning this application or proceeding.
The time period for reply, if any, is set in the attached communication.
Notice of the Office communication was sent electronically on above-indicated "Notification Date" to the
following e-mail address(es):
Patents_eOfficeAction@WolfGreenfield.com
WGS_eOfficeAction@WolfGreenfield.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte STEFAN BASSARAB, BARBARA ENENKEL,
PATRICK GARIDEL, HEIDRUN SCHOTT, SANJAYA SINGH, and
TOBIAS LITZENBURGER1

Appeal 2020-000370
Application 13/494,398
Technology Center 1600
Before ERIC B. GRIMES, LINDA M. GAUDETTE, and
FRANCISCO C. PRATS, Administrative Patent Judges.

GRIMES, Administrative Patent Judge.
DECISION ON APPEAL
This is an appeal under 35 U.S.C. § 134(a) involving claims to an
antibody-containing pharmaceutical composition, which have been rejected
as obvious. We have jurisdiction under 35 U.S.C. § 6(b).
We AFFIRM.

1 Appellant identifies the real party in interest as ABGENOMICS
COOPERATIEF U.A. Appeal Br. 3. We use the word “Appellant” to refer
to “applicant” as defined in 37 C.F.R. § 1.42.
Appeal 2020-000370
Application 13/494,398

2
STATEMENT OF THE CASE
P-selectin glycoprotein ligand-1, or PSGL-1, “is likely to contribute to
pathological leukocyte recruitment in many inflammatory disorders . . . ,
suggesting that inhibitors of PSGL-1, such as antibodies to PSGL-1, are
potentially useful anti-inflammatory drugs.” Spec. ¶ 4. One antibody to
PSGL-1, known as h15A7, has been described in the prior art. Id. ¶ 163.
“Antibody h15A7 is an IgG4 isotype antibody.” Id. Appellant introduced a
“Ser228 → Pro228 mutation . . . into the hinge region of h15A7. The resulting
hinge mutant of h15A7 is called 15A7H herein. . . . The remaining protein
sequence of 15A7H is identical to h15A7.” Id. Antibody 15A7H includes all
of the properties recited for the antibody in the claimed composition. Id.
¶¶ 20, 21.
“IgG4 antibodies are known to undergo a proccess [sic] called Fab
arm exchange, also known as IgG4 shuffling, in which . . . the heavy chains
. . . separate and randomly re-associate to produce a mixed population of
IgG4 molecules with randomized heavy-chain and light-chain pairs.” Id.
¶ 31. “It has been demonstrated that a Serine to Proline mutation . . . at
position 228 using the EU index . . . in the hinge region of human IgG4
results in the reduction of IgG4 ‘half-antibody’ molecules and reduced
heterogeneity/shuffling of IgG4 molecules.” Id. ¶ 32. The Specification
discloses that, as a result of “the Ser228 → Pro228 mutation . . . [t]he amount
of half antibody molecules was reduced from ~8–10% for h15A7 to below
1% for 15A7H.” Id. ¶ 167.
Claims 7, 8, 10–12, and 22 are on appeal. Claim 7, reproduced below,
is illustrative:
Appeal 2020-000370
Application 13/494,398

3
7. A pharmaceutical composition comprising
(i) a monoclonal antibody which immunospecifically
binds to human PSGL-1 comprising:
(a) a variable light (“VL”) chain region comprising
the amino acid sequence of SEQ ID NO: 3;
(b) a heavy chain comprising variable heavy
(“VH”) chain region comprising the amino acid
sequence of SEQ ID NO: 4; and
(c) a human IgG4 constant region which contains a
Serine to Proline substitution at amino acid 228 of
the heavy chain numbered according to the EU
index; and
(ii) a pharmaceutically acceptable carrier,
wherein, the pharmaceutical composition comprises less than
1% half antibody molecules.
OPINION
Claims 7, 8, 10–12, and 22 stand rejected under 35 U.S.C. § 103(a) as
obvious based on Lin,2 Stubenrauch,3 Mataraza,4 Labrijn ’366,5 Osorio,6
Graus,7 and Demarest.8 Final Action9 15. The Examiner finds that Lin
teaches “anti-PSGL antibodies and compositions thereof, including the 15A7

2 Lin et al., US 2009/0191204 A1, published July 30, 2009.
3 Kay Stubenrauch et al., “Impact of Molecular Processing in the Hinge
Region of Therapeutic IgG4 Antibodies on Disposition Profiles in
Cynomolgus Monkeys,” Drug Metabolism and Disposition 38(1):84–91
(2010).
4 Mataraza et al., US 2012/0183565 A1, published July 19, 2012.
5 Labrijn et al., US 2011/0086366 A1, published Apr. 14, 2011.
6 Osorio et al., US 2010/0233157 A1, published Sept. 16, 2010.
7 Graus et al., US 7,563,441 B2, issued July 21, 2009.
8 Demarest et al., US 2009/0092614 A1, published Apr. 9, 2009.
9 Office Action mailed April 16, 2018.
Appeal 2020-000370
Application 13/494,398

4
antibody . . . for a variety of utilities.” Id. at 16. The Examiner finds that Lin
does not describe the serine-to-proline substitution (S228P) recited in
Appellant’s claims but finds that this substitution was known in the art based
on Stubenrauch, Mataraza, Labrijn ’366, Osorio, and Graus.
In particular, the Examiner finds that Stubenrauch “concluded that . . .
IgG4 swapping in vivo was markedly attenuated by S228P mutation in
efforts to stabilize therapeutic IgG4 antibodies with a reduced or eliminated
of [sic] in vivo Fab arm exchange.” Id. The Examiner also finds that Labrijn
’366 teaches that “replacement of the core-hinge region resid[u]e Ser228 by
Pro (S228P) results in partial stabilization of an IgG4 molecule in vitro and
in vivo.” Id.
The Examiner concludes that,
[g]iven the availability of the anti-PSGL-1 antibody 15A7
taught by Lin . . . and the teachings . . . to reduce in vivo Fab
arm exchange and stabilize therapeutic IgG4 antibodies, one of
ordinary skill in the art at the time the invention was made
would have been motivated to provide anti-PSGL-1 antibodies,
including the 15A7 antibody, with the known modification of
human therapeutic antibodies having the human IgG4 constant
region which contains a serine to proline substitution at amino
acid residue 228.
Id. at 17.
Appellant contends that
a person of ordinary skill in the art would not have recognized
that a pharmaceutical composition comprising a specific anti-
PSGL-1 antibody (e.g., 15A7) that includes an IgG4 heavy
chain comprising the S228P mutation would predictably result
in the composition having the specific property (e.g., containing
less than 1% half antibody molecules) recited in the claims on
appeal.
Appeal 2020-000370
Application 13/494,398

5
Appeal Br. 6. Appellant cites references that, it contends, provide evidence
that the effect of an S228P mutation on an antibody is unpredictable. Id. at
7–9. Appellant also contends that “[t]he extent of the reduction in Fab arm
exchange exhibited by the claimed invention is unexpected, given the
unpredictability in the art.” Id. at 10.
We have considered the evidence cited by Appellant and by the
Examiner and conclude that a preponderance of the evidence supports a
conclusion of obviousness. Lin states that its “invention relates to antibodies
and their derivatives that induce apoptosis upon binding to P-Selectin
Glycoprotein Ligand-1 (PSGL-1) on activated T cells.” Lin ¶ 3. Lin states
that its antibodies are useful for treating subjects “having a condition related
to an excessive or unwanted T cell-mediated immune response, e.g., patients
suffering from autoimmune diseases, transplant rejection, allergic diseases,
or T cell-derived cancers.” Id. ¶ 25. Lin discloses therapeutic compositions
comprising its antibodies and states that, “[g]enerally, the antibody is
suspended in a pharmaceutically-acceptable carrier (e.g., physiological
saline).” Id. ¶ 28.
One anti-PSGL-1 antibody disclosed by Lin is “a humanized 15A7
antibody.” Id. ¶ 6. Appellant’s Specification states that “[a]ntibody h15A7
contains the humanized 15A7 heavy and light chain variable regions.” Spec.
¶ 163. The Specification also states that, other than the recited S228P
mutation, “[t]he remaining protein sequence of 15A7H is identical to
h15A7.” Id. Appellant does not dispute that Lin’s humanized 15A7 antibody
comprises the SEQ ID NO: 3 and SEQ ID NO: 4 regions recited in claim 7.
See Appeal Br. 6.
Appeal 2020-000370
Application 13/494,398

6
Rather, the critical issue in this case is whether the effect of mutating
serine to proline at EU index position 228 (S228P) would have been
predictable. More specifically, the issue is whether a person of ordinary skill
in the art would have reasonably expected the S228P mutation to result in an
antibody-containing pharmaceutical composition comprising less than 1%
half antibody molecules, as required by claim 7.
On that issue, Stubenrauch addresses “whether stabilization of the
core hinge of a therapeutic IgG4 antibody by mutation of Ser228 to Pro
(S228P) would be sufficient to prevent in vivo Fab arm exchange.”
Stubenrauch 84, Abstract. “The results indicated that S228P mutation did
not completely prevent Fab arm exchange in vitro in buffer under reducing
conditions relative to IgG4 WT.” Id. In the “[r]esults of the in vivo studies
. . . [i]n contrast, serum from cynomolgus monkeys dosed with the IgG4
mutant was virtually free of swapped IgG4.” Id. Stubenrauch concludes that
“IgG4 swapping in vivo was markedly attenuated by S228P mutation.” Id.
Stubenrauch’s data address the effect of the S228P mutation on Fab
arm exchange, which can result from the generation of half-antibodies. See
id. at 84, right col. However, Stubenrauch does not disclose data regarding
the effect of the S228 mutation on the presence of half-antibodies,
specifically.
Labrijn ’366 also discusses half-antibody (or “half-molecule”)
formation and Fab arm exchange in IgG4 antibodies. Labrijn ’366 ¶ 3.
Labrijn ’366 states that “[n]atalizumab . . . and gemtuzumab . . . are two
humanized IgG4 antibodies currently approved for human use.” Id. ¶ 4.
Appeal 2020-000370
Application 13/494,398

7
Labrijn ’366 discloses results based on commercial formulations of the two
antibodies. Id. ¶ 107.
Specifically, “[t]o determine the type of core-hinge[,] samples . . . of
natalizumab and gemtuzumab were analyzed” and compared to three
antibodies (IgG1, IgG4 (i.e., WT) and IgG4S228P) directed against EGFR
and an IgG4 (WT) antibody directed against CD20. Id. ¶ 128. Labrijn ’366
discloses that “the IgG4 molecules revealed substantial amounts of half-
molecules in addition to intact antibodies.” Id. However, “[t]he S228P
mutation (IgG4S228P) stabilized the IgG4 molecule as demonstrated by the
loss of half-molecules.” Id.
Labrijn ’366 states that “[a]nalysis of natalizumab revealed the
presence of half-molecules indicative of a wild-type IgG4 core-hinge.
Gemtuzumab, however, showed no half-molecules, indicating a stabilized
core-hinge.” Id. Labrijn ’366 also discloses “the hinge region amino acid
sequences for natalizumab and gemtuzumab.” Id. The disclosed sequences
are reproduced below (differing amino acid emphasized):
Natalizumab: YGPPCPSCPAPEFLGGPSVFLFPPKPK
Gemtuzumab: YGPPCPPCPAPEFLGGPSVFLFPPKPK
The only amino acid difference, thus, is an S (serine) in natalizumab (WT
core-hinge) versus a P (proline) in gemtuzumab (stabilized core-hinge).
In summary, Labrijn ’366 discloses that gemtuzumab, with a
stabilized core-hinge having the same S228P mutation as IgG4S228P,
“showed no half-molecules” in the pharmaceutical composition obtained
from the manufacturer.
Labrijn ’366 also discloses further data “suggesting that IgG4 core-
hinge stabilization prevents Fab-arm exchange in vitro.” Id. ¶ 129. Finally,
Appeal 2020-000370
Application 13/494,398

8
Labrijn ’366 discloses that its data showed that, for “hinge-stabilized IgG4
(IgG4S228P-EGFR and gemtuzumab) . . . core-hinge stabilization prevented
IgG4 Fab-arm exchange in vivo, although we can not rule out that low-level
exchange below the level of detection (<0.5% in 72 hrs) of hinge-stabilized
IgG4 does occur.” Id. ¶ 130.10
Labrijn ’366 thus provides evidence that the S228P mutation
eliminated half-antibody generation in compositions comprising
IgG4S228P-EGFR (“the loss of half-molecules”) or gemtuzumab (“no half-
molecules”). This disclosure would have provided a person of ordinary skill
in the art with a reasonable expectation that half-antibody generation would
be eliminated by modifying the humanized 15A7 antibody disclosed by Lin
to include the S228P mutation, and suspending the modified antibody in
physiological saline (Lin ¶ 28).
Lin and Labrijn ’366 provide a reason to modify Lin’s humanized
15A7 to include the S228P mutation, because Lin discloses that
pharmaceutical compositions comprising its antibody are useful in treating
diseases (Lin ¶¶ 25–28) and Labrijn ’366 teaches that the Fab arm exchange
resulting from half-antibody formation can cause adverse effects when
antibodies are used therapeutically (Labrijn ’366 ¶¶ 3, 6–7). Thus, it would

10 The Examiner also cites Mataraza, Osorio, and Graus with respect to the
S228P mutation. Final Action 16–17. These references discuss the S228P
mutation in general terms, but they do not shed light on its effect on half-
antibody generation. See Mataraza ¶ 396; Osorio ¶ 161; Graus 11:16–23.
Therefore, they do not add any disclosure that is material to the rejection on
appeal.
Appeal 2020-000370
Application 13/494,398

9
have been obvious to modify humanized 15A7 to include the S228P
mutation to avoid the adverse effects discussed by Labrijn ’366.
However, Appellant argues that “[a]t the time of the claimed
invention, there was unpredictability in the art with regard to antibody
stabilization in pharmaceutical compositions.” Appeal Br. 7. Appellant cites
Wang11 as teaching that “[a]lthough antibodies share certain structural
similarities, development of commercially viable antibody pharmaceuticals
has not been straightforward because of their unique and somewhat
unpredictable solution behavior.” Id. (citing Wang 1, abstract). Appellant
argues that, based on this disclosure, a skilled artisan “would have
understood that certain modifications that were sufficient to stabilize one
antibody were not necessarily sufficient to stabilize another antibody having
a different amino acid sequence.” Id.
Appellant has not, however, pointed to any disclosure in Wang that
addresses the specific issue relevant to this appeal: the effect of the S228P
mutation on half-antibody generation in a pharmaceutical composition.
Wang therefore carries no weight as evidence on the dispositive issue here.
Appellant argues, though, that “[t]here was also unpredictability in the
art with respect to the effects of the S228P mutation on antibody stability.”
Appeal Br. 7. Appellant points to the statement in Labrijn (2009)12 that
“[t]hus, core-hinge stabilization alone prevents IgG4 Fab-arm exchange in

11 Wang et al., “Antibody Structure, Instability, and Formulation,” J. Pharm.
Sci. 96:1–26 (2007).
12 Labrijn et al., “Therapeutic IgG4 antibodies engage in Fab-arm exchange
with endogenous human IgG4 in vivo,” Nature Biotechnol. 27:767–771
(2009).
Appeal 2020-000370
Application 13/494,398

10
vivo, although we cannot eliminate the possibility of low-level exchange of
hinge-stabilized IgG4 below the level of detection (<8.3% in 72 h;
Supplementary Fig. 2).” Id. (citing Labrijn (2009) at 768, right col.). In
contrast, Appellant argues, Schuurman13 “discloses ‘half molecule exchange
occurs between an IgG4 molecule containing an IgG1 hinge [e.g., having the
S228P mutation] and IgG4 wt molecules.’” Id. (citing Schuurman 63:6–9,
bracketed material in original). Appellant argues that a skilled artisan would
understand “arm exchange,” as used by Schuurman, to mean “half-molecule
exchange in excess of 5%.” Id. (citing Schuurman 63:9–10).
Appellant concludes that
[i]n view of the conflicting teachings of Labrijn (2009) and
Shuurman [sic], a person of ordinary skill in the art at the time
the claimed invention was made would not have been able to
reasonably predict whether the IgG4 S228P mutation, or any
other mutation, would result in reduction of half-antibody
molecules to an amount below 1% in a pharmaceutical
composition comprising any antibody, let alone the specific
anti-PSGL-1 antibody of the claimed pharmaceutical
compositions.
Id. at 8.
We do not find the cited disclosures to be persuasive evidence of a
lack of predictability with respect to the relevant issue: whether a
pharmaceutical composition comprising the antibody recited in claim 7
would comprise less than 1% half antibody molecules. The disclosure in
Schuurman that Appellant points to reports results obtained “under [certain]
in vitro conditions (0.1 mM GSH).” Schuurman 63:4. “GSH” is reduced
glutathione. Id. at 54:11–12. Schuurman states that the GSH was added to

13 WO 2008/119353 A1, published October 9, 2008.
Appeal 2020-000370
Application 13/494,398

11
“restore” (i.e., promote) “exchange activity.” See id. at 55:12–17. These
conditions reflect Schuurman’s goal: to provide “reducing conditions” so
that “two IgG4-or IgG4-like antibodies having different antigen-binding
specificities can perform highly efficient half-molecule exchange and thus
form bispecific antibodies.” Id. at 2:11–13. Appellant has not shown that the
conditions of the cited experiment represent those of a “pharmaceutical
composition” comprising an antibody and a “pharmaceutically acceptable
carrier,” as claimed.
Similarly, the disclosure in Labrijn (2009) that Appellant points to
does not represent results from a pharmaceutical composition. Rather, those
results were found after antibody compositions were “injected . . . into
severe combined immunodeficient (SCID) mice[,] [b]lood samples were
drawn at different times and bispecific antibodies were quantified.” Labrijn
(2009) at 768, right col.
In addition, the disclosures cited by Appellant in both Schuurman and
Labrijn (2009) do not include data that quantify the amount of half-
antibodies, but rather the amount of arm exchange that happens under in
vitro or in vivo conditions. As Appellant’s Specification explains, in IgG4
antibodies, “susceptibility of native IgG4 hinge disulfide bonds to reduction
allows the heavy chains to separate.” Spec. ¶ 31. The result of this separation
of an IgG4 antibody is two half-antibodies. Claim 7 recites that the claimed
composition comprises no more than 1% half-antibodies.
Half-antibodies from different original antibodies can “randomly re-
associate to produce a mixed population of IgG4 molecules with randomized
heavy-chain and light-chain pairs.” Spec. ¶ 31. This process is “called Fab
Appeal 2020-000370
Application 13/494,398

12
arm exchange.” Id. Thus, the levels of arm exchange that Appellant points to
in Schuurman and Labrijn (2009) are evidence that half-antibodies have
recombined (under the in vitro or in vivo conditions used) but the amount of
the resulting mixed (bispecific) antibodies is not the same as the amount of
half-antibodies present.
In fact, when Labrijn (2009) addresses the quantity of half-antibodies
(or half-molecules) in a commercial composition of gemtuzumab, it reports
the same results as Labrijn ’366: “gemtuzumab showed no half-molecules
(indicating a stabilized core-hinge).” Labrijn (2009) at 767, right col. See
also id. at “ONLINE METHODS” (gemtuzumab obtained from
manufacturer).
In summary, we do not find Appellant’s evidence—relating to
quantification of a different compound than recited in claim 7, in
compositions other than a pharmaceutical composition—to be persuasive of
a lack of expectation of success based on the references cited by the
Examiner.
Appellant also argues that van de Winkel14 “discloses that mutations
within the CH3 domain in human IgG4 could prevent Fab arm exchange,
and thus stabilize IgG4, even in the absence of the Ser to Pro mutation of the
core hinge region sequence.” Appeal Br. 9 (citing van de Winkel ¶ 11).15
Appellant argues that “[b]ased upon these teachings, a person of ordinary

14 US 2010/0267934 A1, published October 21, 2010.
15 Appellant also points to van de Winkel’s paragraph 228 and the
accompanying table (Appeal Br. 9), but the disclosures therein are identical
to those cited by Appellant in Schuurman, which have already been
discussed.
Appeal 2020-000370
Application 13/494,398

13
skill in the art would understand that more than one antibody-stabilizing
mutation exists and that such mutations do not always result in inhibition of
Fab arm exchange.” Id.
We do not agree with Appellant’s reasoning. The fact that other
mutations also have the effect of stabilizing an IgG4 antibody (i.e., reducing
or preventing half-antibody formation) is not evidence that the S228P
mutation does not have that effect. Thus, van de Winkel does not cast doubt
on the evidence of record stating that a pharmaceutical composition of
gemtuzumab—with an S228P mutation—“showed no half-molecules.”
Labrijn ’366 ¶ 128.
Finally, Appellant asserts that it “has provided evidence in the form of
experimental data indicating that the claimed pharmaceutical compositions
comprise less than 1% half antibody molecules, retain PSGL-1 binding
activity, and are effective in vivo.” Appeal Br. 10 (citing Spec. ¶¶ 167, 177,
189; Table 7). Appellant argues that “[t]he extent of the reduction in Fab
arm exchange exhibited by the claimed invention is unexpected, given the
unpredictability in the art.” Id.
We have considered the cited evidence, but do not find it persuasive
of nonobviousness. Appellant points to evidence that h15A7 (with a wild-
type hinge) and 15A7H (with an S228P mutation) “demonstrated
comparable binding to activated T cells).” Spec. ¶ 177. Appellant also points
to evidence that “both h15A7 and 15A7H inhibited the trans vivo DTH in a
dose dependent manner.” Id. ¶ 189. “DTH” is delayed-type hypersensitivity.
Id. ¶ 178. Appellant has not explained how either of these results support the
allegedly unexpected reduction in Fab arm exchange.
Appeal 2020-000370
Application 13/494,398

14
Appellant also points to the Specification’s statement that “the
Ser228 → Pro228 mutation significantly lowered the formation of the intra-
chain disulfide bond in the hinge region. The amount of half antibody
molecules was reduced from ~8–10% for h15A7 to below 1% for 15A7H.”
Spec. ¶ 167.
While this evidence does relate to Appellant’s assertion of unexpected
results, Appellant has not adequately shown that it would, in fact, have been
unexpected by a person of ordinary skill in the art. The prior art of record
shows that a pharmaceutical composition of a different antibody having the
same Ser228 → Pro228 mutation “showed no half-molecules.” Labrijn ’366
¶ 128. In view of that knowledge in the art, Appellant has not persuasively
shown that the Specification’s evidence of less than 1% half antibody
molecules for the 15A7H antibody, also with the Ser228 → Pro228 mutation,
would not have been expected.
In summary, a preponderance of the evidence of record supports a
conclusion of obviousness with respect to claim 7. The rejection of claim 7
under 35 U.S.C. § 103(a) is affirmed. Claims 8, 10–12, and 22 were not
argued separately and therefore fall with claim 7. 37 C.F.R. § 41.37(c)(1)(iv).
DECISION SUMMARY
In summary:
Claims
Rejected
35 U.S.C. § Reference(s)/Basis Affirmed Reversed
7, 8, 10–12,
22
103(a) Lin, Stubenrauch,
Mataraza, Labrijn ’366,
Osorio, Graus,
Demarest
7, 8, 10–
12, 22

Appeal 2020-000370
Application 13/494,398

15
TIME PERIOD FOR RESPONSE
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a). See 37 C.F.R.
§ 1.136(a)(1)(iv).
AFFIRMED