`Patent No. 8,729,094
`Petition for Inter Partes Review
`Attorney Docket No. REDDY 7.1R-010
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`DR. REDDY'S LABORATORIES, LTD. and
`DR. REDDY'S LABORATORIES, INC.
`Requestors
`
`v.
`
`HELSINN HEALTHCARE S.A. and ROCHE PALO ALTO LLC
`Patent Owner
`
`Patent No. 8,729,094
`Issue Date: May 20, 2014
`Title: LIQUID PHARMACEUTICAL FORMULATIONS OF PALONOSETRON
`
`Inter Partes Review No. Unassigned
`
`DECLARATION OF DAVID G. FRAME. PHARM.D.
`
`(Exhibit 1023)
`
`4117631_1.docx
`
`Dr. Reddy's Laboratories, Ltd., et al.
`v.
`Helsinn Heallhcare S.A., et al.
`U.S. Patent No. 8,729,094
`Reddy Exhibit 1023
`
`Exh. 1023
`
`
`
`I, DAVID G. FRAME, hereby declare as follows:
`
`1.
`
`2.
`
`I am a US citizen and a resident of the state of Michigan.
`
`I received a B.S. in Chemistry from St. Louis University in 1986, and
`
`a B.S. in Pharmacy and Pharm. D. in 1993 and 1994, respectively, from Wayne
`
`State University.
`
`3.
`
`I am presently Assistant Professor of Pharmacy at the University of
`
`Michigan and Hematology/Oncology/BMT Clinical Specialist with the University
`
`of Michigan Health System. In my present positions, I teach therapeutics and
`
`pharmacology at the University of Michigan, College of Pharmacy, and, in
`
`conjunction with physicians, I treat patients who are receiving chemotherapy for
`
`hematologic malignancies and for those receiving a bone marrow transplant. In my
`
`work treating patients, I have established our antiemetic guidelines and do much of
`
`the supportive care management of these patients. I also do both clinical and
`
`translational research in the fields of oncology and infectious diseases.
`
`4.
`
`Prior to joining the University of Michigan, I was Assistant Professor
`
`at Rush University from 1995-2006. While working at Rush University, I chaired
`
`the Chemotherapy Overview Committee at Rush Medical Center from 1995 to
`
`2005. I was Director of Clinical Hematology/Oncology Pharmacy Services and
`
`Research at Rush University Medical Center from 1999 to 2005. In that role, I was
`
`Exh. 1023
`
`
`
`responsible for overseeing all clinical pharmacy oncology services and monitoring
`
`oncology patient outcomes. I was also director of the anticoagulation program for
`
`hip and knee replacement where we were responsible for supportive care issues,
`
`such as nausea and vomiting post-operatively.
`
`5.
`
`I have taught graduate courses and have given frequent presentations
`
`on many topics in oncology, including therapies for multiple disease conditions
`
`and lectures on supportive care. I have delivered over 50 professional presentations
`
`to clinicians on the treatment of both chemotherapy induced and post-operative
`
`nausea and vomiting.
`
`6.
`
`I have participated m designing clinical studies of 5-HT3 receptor
`
`antagonists for treating both chemotherapy induced nausea and vomiting and post-
`
`operative nausea and vomiting. My role in the clinical studies has involved being a
`
`part of large multicenter trials, as well as designing and implementing my own
`
`randomized controlled trial for post-operative nausea and vomiting, as well as
`
`several smaller trials in the oncology setting.
`
`7.
`
`In my years of clinical practice, I have had experience with all of the
`
`5-HT3 antagonists approved in the U.S., including Aloxi® (palonosetron), Zofran®
`
`(ondansetron), Kytril® (granisetron), and Anzemet® (dolasetron). I have managed
`
`nausea and vomiting using these agents in over 4000 patients.
`
`2
`
`Exh. 1023
`
`
`
`8.
`
`Additional details of my education and experience are set forth in my
`
`curriculum vitae (Exh. 1024 ), which also contains a partial listing of the companies
`
`and conferences where I have been invited to speak, the lectures and short courses
`
`I have given, and the publications I have authored.
`
`9.
`
`I was retained by litigation counsel for Dr. Reddy's Laboratories, Ltd.
`
`and Dr. Reddy's Laboratories, Inc. to serve as an expert and to testify in some of
`
`the litigations between Helsinn Healthcare S.A. and Roche Palo Alto LLC and
`
`Dr. Reddy's Laboratories, Ltd., Dr. Reddy's Laboratories, Inc., identified to the
`
`Patent Trial and Appeal Board in the Notice of Related Matters in the Petition that
`
`this declaration supports.
`
`10.
`
`I have been separately retained by Lerner, David, Litten berg,
`
`Krumholz & Mentlik, LLP ("counsel") to provide my opinions in this Inter Partes
`
`Review in the fields of pharmaceutical design and administration, the structure and
`
`analysis of clinical studies and their data and specifically the formulation and
`
`administering of palonosetron and other 5-HT3 receptor antagonists recognized as
`
`having structural or functional relationships with palonosetron (sometimes
`
`collectively referred to as "setrons"). I have read and understood U.S. Patent
`
`No. 8,729,094 ("the '094 Patent") (Exh. 1001), as well as all other references
`
`discussed in this declaration. I am being compensated for my time in an amount
`
`3
`
`Exh. 1023
`
`
`
`consistent with my customary consulting fee, and my compensation IS not
`
`contingent on my opinion or the outcome of this proceeding.
`
`I.
`
`A PERSON OF ORDINARY SKILL IN THE ART
`11.
`I understand from counsel that patents such as the '094 Patent are
`
`neither addressed to experts nor to laymen; rather they are addressed to persons of
`
`ordinary skill in the relevant art at the time invention was made, which I have been
`
`told by counsel to assume is January 29, 2003. I also understand from counsel that
`
`factors relevant to the level of skill in the art include, without limitation: the
`
`educational level of the inventor, the types of problems encountered in the relevant
`
`area, prior art solutions to those problems, the rapidity with which innovations are
`
`made, the sophistication of the technology, and the educational level of active
`
`workers in the field.
`
`12. As noted previously, I have attended and testified at the Trial which
`
`involved other patents related to the ' 094 Patent. I am not an expert in patent law
`
`but, from my recollection, none of those patents involves methods of treating
`
`CINV per se. That must be contrasted with claims 22-30 of the '094 Patent, which
`
`appear on their face to relate to methods of treating CINV, albeit by administering
`
`the types of formulation that are the focus of the above-identified litigations.
`
`13. Because of my involvement in the aforementioned litigations, I know
`
`that Petitioner has taken the position that a person of ordinary skill in the art (a
`
`4
`
`Exh. 1023
`
`
`
`"POSA") as to the patents at issue in those cases, would be a formulation scientist,
`
`typically with a PhD in pharmaceutics or a related field, and would have a couple
`
`of years experience in developing IV formulations and bench experience. I also
`
`understand that Petitioner argued that formulation scientists would draw on the
`
`pharmaceutical science literature, general text books, research articles and
`
`abstracts, and other sources of information, including from clinicians and
`
`pharmacologists, and other scientists in the field.
`
`14.
`
`I would certainly agree that this definition is still broadly applicable
`
`here; and as regards the aspects of the claimed invention that relate to drug
`
`formulation and dose, I consider myself to be a person of greater than ordinary
`
`skill. However, the claimed invention in the '094 Patent is, in my view, primarily
`
`directed to a method of medical treatment; and as such, I believe medical doctors
`
`and other medical professionals having experience specifically treating cancer and
`
`addressing the side effects of the use of emetogenic chemotherapies would be of at
`
`least equal import. Such doctors or clinicians would have a high level of formal
`
`education and several years of practical experience. In particular, they should have
`
`experience in clinical study, and an interpretation of data from medical studies as
`
`would be applied to treating patients. Whether these medical and claimed
`
`professionals would consult with formulators or the reverse, is of lesser importance
`
`to me, as all of their collectible knowledge and experience are implicated by the
`
`5
`
`Exh. 1023
`
`
`
`'094 Patent. I understand from counsel that a POSA may be a composite of
`
`different types of individuals, with each individual having a somewhat different
`
`background from the others. When it comes to the claims of the '094 Patent, in my
`
`opinion, a POSA would most appropriately be a composite of the types of
`
`individuals I have just noted.
`
`II. THE '094 PATENT (EXHIBIT 1001)
`15. Reading the '094 Patent, surprisingly little information is provided
`
`regarding methods of medical treatment. Most of the specification, which I
`
`understand this patent shares with several prior patents, is directed to stabilized
`
`formulations. But the claims talk of a method of treating chemotherapy induced
`
`nausea and vomiting (' CINV"). In its Background section, the
`
`' 094 Patent
`
`acknowledges a prior U.S. patent, Berger, U.S. Patent No. 5,202,333 (Exh. 1010),
`
`and that clinical investigations had been made concerning palonosetron. "These
`
`investigations have shown that the drug is an order of magnitude more potent than
`
`most 5-HT3 receptor antagonists, has a surprising half-life of about 40 hours, and is
`
`effective to reduce delayed-onset nausea induced by chemotherapeutic agents."
`
`(Exh. 1001 col.1 11.57-62.)
`
`16. The '094 Patent also acknowledges in its Background section that
`
`palonosetron formulations, and, in particular, an IV formulation used to treat
`
`CINV, were known. However, characterizing the IV formulation of Example 13 of
`
`6
`
`Exh. 1023
`
`
`
`Berger, the ' 094 Patent notes that "[t]he formulation has a pH of 3.7 and a shelf
`
`stability of less than the 1-2 year time period required by health authorities in
`
`various countries." (Exh. 1001 col.2 11.1 0-12.) This obviously is a stability issue,
`
`not an issue that impacts patients directly. Moreover, in the U.S., manufacturers
`
`can set their own "pull date," meaning the date that the product has to be removed
`
`from the market. (See Exh. 1020.) They must provide stability testing to support
`
`their shelf-life claim, but the fact that a product did not have a shelf life of one to
`
`two years would not be a regulatory hurdle in the United States. Ironically, as the
`
`Wall Street Journal article (Exh. 1 020) explains, drug companies would generally
`
`prefer the "pull date" to be shorter to increase sales.
`
`17. Notwithstanding the alleged stability Issues facing
`
`the Berger
`
`formulation, a POSA looking at palonosetron at the relevant time would know of
`
`palonosetron' s potential real world chances of reaching the commercial market.
`
`Preclinical, Phase I and Phase II clinical study data had been made public, and all
`
`of that published data established safety and efficacy for palonosetron over a range
`
`of doses tested. (See Exhs. 1010-1012, 1018, 1019.) Also important to a POSA's
`
`understanding at the time are the various press releases made by Patent Owner,
`
`such as the press releases I have reviewed from April 10, 2001, October 3, 2001 ,
`
`and January 16, 2002. (Exhs. 1033-1035.) Looking at the January 16, 2002 press
`
`release (Exh. 1 033), as an example, Patent Owner announced that "palonosetron is
`
`7
`
`Exh. 1023
`
`
`
`a potent, highly selective 5-HTrreceptor antagonist m development in North
`
`America and Europe for the prevention of chemotherapy induced nausea and
`
`vomiting (CINV)." (Exh. 1 033.) Moreover, submission of a New Drug Application
`
`(''NDA") for palonosetron was planned for the third quarter of 2002. Still further,
`
`Patent Owner announced that a "Phase 3 clinical trial program" had been initiated
`
`in April 2000, and that Patent Owner was "pleased to have completed all patient
`
`treatment and to have begun analysis ofthe data." (Jd.)
`
`18. Because of this public information, a POSA would have more than
`
`merely a reasonable expectation that IV palonosetron would be sufficiently
`
`successful in treating CINV to be Food and Drug Administration ("FDA")
`
`approved. Indeed, the FDA would not have granted permission for Patent Owner to
`
`take this product to Phase III, and Patent Owner would not announce its intention
`
`to file its new drug application, unless both had a strong expectation of success.
`
`19. Patent Owner also explained in its press release that it expected, based
`
`on the extended half-life of palonosetron, and the results of prior studies, that
`
`palonosetron would be effective in treating delayed emesis, and it was testing that
`
`expectation in Phase III. (See id. ("[t]he activity seen with palonosetron in the
`
`Phase 2 trial, coupled with its safety profile observed to date, led to the initiation of
`
`a Phase 3 program to assess the ability of the drug to provide prolonged protection
`
`against CINV with a single dose.").)
`
`8
`
`Exh. 1023
`
`
`
`20. Thus, by January 29, 2003, a POSA would have information on an
`
`exemplary palonosetron formulation whose stability was at least sufficient for
`
`clinical use. A POSA had reason to expect that a palonosetron IV process for
`
`treating CINV was safe and effective. And a POSA had information on effective
`
`dosing ranges from published Phase II studies. Finally, a POSA would have reason
`
`to believe that palonosetron was more potent than its peers and able to treat
`
`delayed emesis.
`
`Ill. DATA CONVERSION TO UNIFORM UNITS
`21. Claim 22 of the '094 Patent expresses the formulation used in the
`
`claimed method in terms of a number of milliliters of a palonosetron solution at a
`
`certain concentration in mg/ml. Specifically, 5ml of a solution having a
`
`concentration of 0.05mg/ml is claimed, which yields a dose of 0.25mg. Much of
`
`the prior art discussed herein, however, expresses doses in alternate units. This can
`
`be confusing, making it difficult to relate one set of numbers to another.
`
`22. Accordingly, I will employ techniques used by POSAs to convert
`
`everything to a common set of units; namely, milligrams ("mg"). This will allow
`
`doses disclosed in the prior art to be expressed in units that can be directly
`
`compared to the doses claimed. It also allows concentrations to be considered in
`
`the units described in claim 22; namely, mg/ml. For the convenience of the Patent,
`
`Trial and Appeal Board, I have made the necessary calculations for each of the
`
`9
`
`Exh. 1023
`
`
`
`doses in each of the various references I rely upon and converted everything to
`
`equivalent human doses, which are provided in Table 1 below.
`
`TABLE 1
`
`Eglen
`Eglen
`(Human/Ferret) (Human/Dog)
`
`Chelly
`(Human)
`
`Tang
`(Human)
`0.007 mg
`
`0.012 mg
`
`0.021 mg
`
`0.021 mg
`
`0.013 mg
`
`0.04 mg
`
`0.13 mg
`
`0.4mg
`
`1.3 mg
`
`0.04 mg
`
`0.12 mg
`
`0.4mg
`
`1.2mg
`
`4.0mg
`
`12.0 mg
`
`0.07 mg
`
`0.21 mg
`
`0.7mg
`
`2.1 mg
`
`0.07 mg
`
`0.21 mg
`
`0.7mg
`
`2.1 mg
`
`Tested p.tg/kg
`
`0.1
`
`0.3
`
`1
`
`3
`
`10
`
`30
`
`100
`
`300
`
`Also for the Board's convenience, I will recite the units for dosing in the actual
`
`units used in each reference. However, I will follow that, in each instance, with the
`
`dose converted to mg for humans, which will follow in [square brackets].
`
`23.
`
`Specifically, animal data, and indeed human data, is often reported as
`
`an amount of the active ingredient given per kg of the patient's body weight (e.g.,
`
`3.0J..Lg/kg). Traditionally, when pharmacologists and medical professionals need to
`
`convert such data from fJ.g/kg to, for example, mg, they first convert to mg/kg and
`
`10
`
`Exh. 1023
`
`
`
`then multiply by a "standard" 70kg individual - about 155lb. This number is an
`
`approximate average of the weight of men and women and provides a standardized
`
`basis for discussion.
`
`24.
`
`I have not reviewed the entire prosecution history of the '094 Patent,
`
`which I understand to be a written record of the communications between the U.S.
`
`Patent and Trademark Office and the Patent Owner seeking the '094 Patent.
`
`However, I have reviewed applicants' response of February 21, 2014 (Exh. 1009),
`
`in which the Patent Owner cited an Examiner's prior statement made in connection
`
`with U.S. Patent No. 8,958,219 (Exh. 1005). I understand this '219 Patent to be the
`
`'094 Patent's immediate predecessor. In the prosecution of the '219 Patent, the
`
`Examiner relied upon this same 70kg standard weight conversion in explaining his
`
`position. (See Exh. 1009, at 8-9.) Berger owned by Patent Owner, used this
`
`conversion as well. (See Exh. 1010 col.12 11.16-18 ("Therefore, a therapeutically
`
`effective amount for a 70kg human may range from .... ").) Using this conversion,
`
`a dose of 1 f.!g/kg would be converted to a standard measure for humans in mg by
`
`converting to 0.001mg/kg and multiplying by 70kg, for a total of 0.07mg. A dose
`
`of30f.lg/kg would convert to a dose of2.lmg.
`
`25. For animal data, an additional factor must be used in the computation
`
`to arrive at a comparable human dose in mg. More specifically, to estimate
`
`equivalent doses in humans from animal doses, the animal doses are adjusted based
`
`11
`
`Exh. 1023
`
`
`
`on a body surface area normalization usmg a common conversiOn factor as
`
`described in the literature. (See Exhs. 1043; 1047.) The extrapolation of the animal
`
`dose to human dose is calculated by dividing the dose in each of the animal species
`
`studied by the appropriate body surface area conversion factor ("BSA-CF"). The
`
`conversion factor is a unitless number that converts an J!glkg dose for each animal
`
`species to an J!glkg dose in humans. That would in turn be converted from J.lglkg to
`
`mg/kg and next would be multiplied by the standard 70kg human. For ferrets, the
`
`animal dose is divided by a factor of 5.3. (See Exh. 1031, at 6-7 (Table 1 ).) Thus, a
`
`dose of 30f.lg/kg administered to a ferret as reported in Eglen would first be
`
`converted to milligrams (0.03mg/kg), divided by 5.3 and then multiplied by the
`
`70kg number discussed immediately above for a standard human resulting in an
`
`equivalent dose of 0.4mg.
`
`26. The conversion factor for dogs is 1.8. (See Exh. 1031 at 6-7
`
`(Table 1).) However, the process is otherwise identical. Thus, the equivalent
`
`dosage for dogs of a dose of, for example, 3 f.lg/kg would involve converting that
`
`dose to 0.003mg/kg, dividing by 1.8 and multiplying by 70kg for an equivalent
`
`dose of0.12mg.
`
`IV. EGLEN
`27. Eglen was published in 1995 and reports the results of preclinical
`
`antiemetic studies of palonosetron performed in ferrets and dogs. Both models are
`
`12
`
`Exh. 1023
`
`
`
`art recognized for their usefulness in predicting dose and efficacy in humans.
`
`(Exh. 1011, at 861 ("The ferret and the dog are well established animal models of
`
`emesis which respond to cancer chemotherapeutic agents in a manner similar to
`
`that observed in man.") (citations omitted).) Adult male ferrets were given
`
`1-100J..lg/kg
`
`[0.013-1.3mg] of palonosetron by IV 30 minutes prior
`
`to
`
`administration of the chemotherapeutic agent cisplatin. (!d. at 861.) Adult male
`
`dogs were given palonosetron in an amount of0.3-300).lg/kg [0.012mg-12.0mg] by
`
`IV 120 minutes prior to the administration of chemotherapy. Both were also dosed
`
`with ondansetron as a comparator. Dogs were also pretreated with either drug at
`
`various intervals beginning as early as 24 hours prior to the administration of
`
`cisplatin in an experiment design to examine the duration of antiemetic activity.
`
`28. Eg1en's data is best illustrated by reference to Figures 5 and 6.
`
`(Exh. 1011, at 863.) Figure 5, reproduced as Exhibit 1051, presents the data for
`
`ferrets dosed at 1, 3, 10, and 30J..Lg/kg [0.013, 0.04, 0.13, and 0.4mg]. (See
`
`Exh. 1051.)
`
`29. At all of these dose levels, inhibitory effect was observed relative to
`
`the placebo, which is marked in Figure 5 as "Veh." This means that for all four of
`
`the doses tested, palonosetron exhibited efficacy in reducing CINV. Somewhere
`
`within the dose range of between about 1 and about lO).lg/kg [0.013-0.13mg], a
`
`plateau was reached. Before that point, the effect of palonosetron on CINV
`
`13
`
`Exh. 1023
`
`
`
`increased with an increasing dose. But thereafter, increasing the dose had little
`
`additional effect. As a consequence, dosing below about 1 J..Lg/kg [0.0 13mg] would
`
`likely be suboptimal. There are not enough data points to know exactly where the
`
`inflection point (where plateau truly flattens out) is located. To be conservative,
`
`therefore, a POSA would view the likely starting doses in humans to be about
`
`0. 04 mg based on ferrets.
`
`30. Generally speaking, in drug development, one would not pick a dose
`
`for further study or commercial use that was only "near" the plateau, favoring
`
`instead a dose firmly on the plateau. However, one also would not pick a dose
`
`relatively far past the start of the plateau, because there would be no additional
`
`benefit to the patient, because it would likely increase the overall cost of the drug,
`
`and because it would likely raise unnecessary formulation or side effect issues. For
`
`humans based on ferrets that dose would likely be approximately 0.4mg.
`
`31. Looking at Figure 6( a) (Exhibit 1 052), the data in dogs was extremely
`
`similar. At a dose of 0.3J..Lglkg [0.012mg], there was almost no difference shown
`
`between palonosetron and placebo. However, an increase in inhibitory activity was
`
`observed at 1 J..Lg/kg [0.04mg], a further increase in activity was observed at 3 and
`
`1 OJ..Lg/kg [0.12mg and 0.4mg], without much gain at larger doses. (See id. at 863
`
`Fig.6.) This would suggest a dose somewhere between about 3 and lOJ..Lg/kg
`
`[0.12mg-0.4mg]. Again, I believe a POSA would consider 3J..Lglkg [0.12mg] a
`
`14
`
`Exh. 1023
`
`
`
`likely human starting dose based on dogs, but would not consider doses as high as
`
`30J.Lg/kg [1.2mg].
`
`32. Eglen also presents m Figure 6(b) comparative data usmg
`
`ondansetron. (Exh. 1011, at 863, Figure 6.) For ondansetron in dogs, inhibiting
`
`activity does not begin until about 30J.Lg/kg [ 4.0mg] for ondansetron in dogs. This
`
`illustrates the superior potency of palonosetron over ondansetron. However, the
`
`data suggests that the same plateau in behavior observed in palonosetron is
`
`common to ondansetron as well. And indeed, the literature supports such a
`
`conclusion in other 5-HT3 receptor antagonists. (Exh. 1027, at 237 ("We conclude
`
`that for each drug there is a plateau in therapeutic efficacy at a definable dose level
`
`above which further dose escalation does not improve outcome.").)
`
`33. Looking at the data from Eglen, I believe a POSA would conclude
`
`from the ferret data, that a likely successful dose would be 3J.Lg/kg [0.04mg] or
`
`more. Based on the dog data, that point would also be about 3J.Lg/kg [0.12mg] or
`
`more. While there is of course a difference between these two species, this data is
`
`relatively close, and the plateaued pattern established by both are very supportive
`
`and in good agreement. In my opinion, Eglen would teach a person of ordinary
`
`skill in the art that palonosetron by IV was effective to treat CINV at reasonably
`
`low doses.
`
`15
`
`Exh. 1023
`
`
`
`34. Eglen also taught the POSA a useful range of concentrations, which
`
`agrees well with the concentration range dictated merely by formulation necessity.
`
`Ferrets were dosed 1-100jlg/kg, and the vehicle control was 1ml/kg. This means
`
`that when the dose of drug was administered it was first dissolved in 1 ml/kg. The
`
`ferrets in this trial were 1-1.4kg, and thus the volume of the solution was 1 to
`
`1.4ml. Accordingly, using a 1kg ferret the dose would have been 1jlg in 1ml (or a
`
`1 J.Lg/ml, which is the same as 0.001 mg/ml solution). As such, the range of
`
`concentrations of palonosetron administered to ferrets was between 0.001 mg/ml
`
`and 0.1 mg/rnl. In dogs, the range was broader. The doses there were 0.30 to
`
`300J.Lg/kg, and the dogs used in this trial were 8-20kg. A vehicle control of
`
`0.1 ml/kg was employed. Thus, the range of concentration was from 0.003mg/rnl to
`
`3mg/ml. Both of these ranges encompass the 0.05mg/ml concentration claimed.
`
`35. Moreover, both ranges match up well with the concentration levels
`
`dictated purely as a consequence of dose selection and the obvious use of a small
`
`volume, which the industry would have demanded. Patent Owner's witnesses at
`
`Trial agreed that the volume used in this instance would be no greater than 5ml.
`
`(Exh. 1032; 1043, at 157:12-21.) I agree this a feasible volume. And the three other
`
`setron products on the market before palonosetron were all sold in single-use,
`
`single-dose forms of 5ml or less. (See Exhs. 1015, at 1503 (2ml single-dose vial);
`
`1016, at 3106 (1 ml single-use vials and 4ml multi-dose vials); 1017, at 682
`
`16
`
`Exh. 1023
`
`
`
`(0.625rnl single-use capsule and 5rnl single-use vial).) So the possibilities in terms
`
`of volume were very limited.
`
`36. Eglen in view of Chelly (as will be discussed below), suggests a
`
`dosing range of between about 1J.1g/kg [0.07mg] and 10J.1g/kg [0.70mg]. Indeed,
`
`for the reasons further discussed, that range is more likely to be between about
`
`3J.1g/kg [0.21mg] and about 10J.1g/kg [0.7mg]. These generally would be
`
`formulated in whole integers of from 1 to 5rnl as noted above. Thus, the lowest
`
`possible concentration of the resulting range would be 0.014mg/ml (0.07mg/5ml),
`
`and the highest would be 0.70mg/ml (0.70mg in a single ml). The more likely
`
`range would be between 0.04mg/rnl (0.21mg/5rnl) through 0.70mg/rnl (0.7mg in a
`
`single rnl). Both ranges clearly encompass the 0.05mg/rnl number claimed.
`
`3 7.
`
`It bears noting that the actual concentration of the drug has no impact
`
`on its clinical efficacy. Indeed, a POSA would know that the clinical efficacy of a
`
`drug comes from the actual dose delivered, not the concentration of the drug in
`
`solution.
`
`V. CHELLY (EXH.1012)
`38. Chelly is an abstract published in September 1996 in the Journal of
`
`Anesthesiology. (Exh. 1012.) Chelly reports Phase II clinical data from humans
`
`where palonosetron was administered orally
`
`to 351 patients
`
`to prevent
`
`postoperative nausea and vomiting ("PONV") from
`
`laparoscopic surgery.
`
`17
`
`Exh. 1023
`
`
`
`Palonosetron was dosed at 0.3, 1.0, 3.0, 10.0, and 30J.!glkg [0.021mg, 0.07mg,
`
`0.21 mg, 0. 7mg, and 2.1 mg, respectively] or placebo. Patients received the study
`
`medication one to two hours prior to surgery. As noted previously, to convert these
`
`doses in Jlg/kg to mg, one first converts to mg/kg and then multiplies by a
`
`"standard" 70kg human. Because the data here is from humans, no additional
`
`conversion is required.
`
`39. As Chelly explained, "[c]ompared to the placebo, RS-25259 increased
`
`significantly the percentage of patients who elicited CR [complete response] (37%,
`
`58%, 52%, 59%, and 53% vs. 33%). RS-25259 therapeutic effectiveness reached a
`
`plateau at a dose of 1J.J.g/kg [0.07mg]. Except for the lowest dose of 0.3Jlg/kg
`
`[0.021mg], RS-25259 induced a significant reduction in the frequency of severe
`
`nausea episodes as compared to placebo." (!d.) This is very aptly illustrated in the
`
`dose/efficacy curve for Chelly presented in Exhibit 1053. Chelly also discussed the
`
`fact that no unexpected adverse reactions occurred. In my opinion, this data
`
`indicates that 0.07mg was the minimum for a possible effective dose for treating
`
`PONY in human patients when palonosetron was administered orally. Once the
`
`actual plateau was reached, as was seen in Eglen's animal models, and from the
`
`literature in other 5-HT3 receptor antagonists (Exh. 1027, at 237), there would be
`
`no
`
`increase in efficacy at a higher dose. Thus, administering additional
`
`palonosetron would provide no greater efficacy. I believe a POSA looking at this
`
`18
`
`Exh. 1023
`
`
`
`data would be conservative because the few data points reported would not permit
`
`one to know the exact location of the inflection point of the plateau. Thus I feel a
`
`POSA would consider 3~g/kg[0.21mg] to be a better starting dose for humans
`
`based on Chelly.
`
`VI. EGLEN IN VIEW OF CHELLY
`40. Because pure receptor antagonists are very predictable in their dose
`
`response properties, it would be expected that the animal data would be
`
`particularly translatable to the human. Thus, it was no surprise to me and, I believe,
`
`would be expected by a POSA, how congruent the data are. Ferrets indicated a
`
`starting dose of about 3 ~g/kg[0.04mg]; and dogs indicated a starting dose of about
`
`3~g/kg[0.12mg]. Again, significant activity was observed at lower doses. But the
`
`number of data points does not permit one to accurately know the exact intlection
`
`point. This is not really an issue as one merely needs to formulate a comfortable
`
`distance from that region to ensure efficacy. Thus I believe a POSA would select
`
`0.04mg and 0.12mg for a starting point for human doses based on the converted
`
`data from ferrets and dogs respectively. This narrow range brackets the data
`
`showing significant activity in humans in Chelly, where the lowest starting doses is
`
`about 1 f.lglkg[0.07mg]. That said, for the same data point reasons, I think a POSA
`
`would likely be inclined to start at 3J.tg/kg[0.21mg] in humans. However, the data
`
`at 1 ~g/kg[0.07mg] in humans from Chelly might suggest a starting dose between
`
`19
`
`Exh. 1023
`
`
`
`1 J.!g/kg[0.07mg] and 3 Jlg/kg[0.21 mg]. And from there, efficacy plateaued in all
`
`three. This is actually excellent agreement. The congruent nature of this data would
`
`provide a POSA with confidence in the validity of the data, and the conclusions
`
`that can be drawn about dosing based on that information.
`
`41 . The starting point for human dosing suggested by the dog and ferret
`
`data align very well with actual human experience. And they all plateau. From my
`
`own experience, I believe that a POSA would not select the very lowest possible
`
`dose based on this data, but would not formulate too far from it. So, while a range
`
`of about 1 Jlg/kg-1 0~-tg/kg [0.07 to 0.7mg] is certainly a possible range based on
`
`human doses in Chelly, I believe a POSA would more likely select a lower dose
`
`closer to 3J.!g/kg [0.21mg].
`
`42.
`
`I find the agreement between the data from these two studies to be
`
`particularly interesting in this case, as Eglen was a study based on IV
`
`administration for CINV, whereas Chelly was based on oral administration for
`
`PONV. Yet the data lined up nicely. Not only is this significant because of its
`
`overall agreement, but for other reasons as well, as I now explain.
`
`4 3. Prior art with other setrons suggests that a higher dose of a particular
`
`5-HT3 receptor antagonist may be necessary when treating CINV rather than
`
`PONV. I estimate that the adjustment for that change in indication could range
`
`from 2x to 1 Ox. I note that Patent Owner's expert at the Trial agreed with this
`
`20
`
`Exh. 1023
`
`
`
`range. (Exh. 1029.) Moreover, dolasetron mesylate is a 5-HT3 receptor antagonist
`
`approved and on the market before palonosetron (see Exh. 1 0 17), and it was
`
`approved for both CINV, and PONY. For CINV, a dose of 100mg was approved as
`
`a single dose. For PONY, however, the dose would be 12.5mg, roughly 1/8th of
`
`the dose used for CINV. (See id. at 683.)
`
`44. Ondansetron hydrochloride was approved in the U.S. at a dose of
`
`32mg for CINV, and 4mg for PONY. This is an 8x differential. (See Exh. 1015,
`
`at 1506.) However, other clinical studies had established that ondansetron doses of
`
`8mg were equally effective in treating CINV. (See Exh. 1044.) An 8mg dose of
`
`ondansetron is a 2x increase over the 4mg dose used for PONV. Granisetron was
`
`also studied for both PONY and CINV. The approved CINV dose was 10J!glkg.
`
`Granisetron had not been approved for PONY at the time, but the doses that were
`
`reported to be effective for PONY were 5J!glkg, again a 2x difference from the
`
`CINV dose. (Exh. 1049.)
`
`45.
`
`Since the setrons are competitive inhibitors, it is clear that the higher
`
`affmity the drug has for the receptor, the smaller will be the decrease in the overall
`
`difference in dosing required as between PONY and CINV. For example,
`
`ondansetron and granisetron had higher affinities than dolasetron and thus needed
`
`only 2x difference in dosing, while dolasetron required 8x difference. Palonosetron
`
`had yet a higher affinity than both ondansetron and granisetron, and thus with
`
`21
`
`Exh. 1023
`
`
`
`palonosetron, it would be anticipated that the PONV and CINV doses,
`
`respectively, could be very similar. (Exh. 1028.)
`
`46. Moreover, Chelly administered palonosetron orally. When a drug is
`
`administered orally, it must ftrst be absorbed from the GI tract, and many