J. Med. Chem. 2007, 50, 6665–6672
`
`6665
`
`Trends in Active Pharmaceutical Ingredient Salt Selection based on Analysis of the Orange
`Book Database
`
`G. Steffen Paulekuhn,†,‡ Jennifer B. Dressman,‡ and Christoph Saal*,†
`
`Merck KGaA, Frankfurter Strasse 250, 64293 Darmstadt, Germany, and Institute of Pharmaceutical Technology, Biocenter, Johann Wolfgang
`Goethe UniVersity, Max Von Laue Street 9, 60438 Frankfurt (Main), Germany
`
`ReceiVed August 20, 2007
`
`The Orange Book database published by the U.S. Drug and Food Administration (FDA) was analyzed for
`the frequency of occurrence of different counterions used for the formation of pharmaceutical salts. The
`data obtained from the present analysis of the Orange Book are compared to reviews of the Cambridge
`Structural Database (CSD) and of the Martindale “The Extra Pharmacopoeia”. As well as showing overall
`distributions of counterion usage, results are broken down into 5-year increments to identify trends in
`counterion selection. Chloride ions continue to be the most frequently utilized anionic counterions for the
`formation of salts as active pharmaceutical ingredients (APIs), while sodium ions are most widely utilized
`for the formation of salts starting from acidic molecules. A strong trend toward a wider variety of counterions
`over the past decade is observed. This trend can be explained by a stronger need to improve physical chemical
`properties of research and development compounds.
`
`physical chemical properties of pharmaceutical salts and meth-
`ods for salt screening exist, e.g., refs 4, 16-19 and references
`included therein. On the other hand, publications giving an
`overview of approved salt forms are very few.1–3 All publica-
`tions known to the authors dealing with occurrence of coun-
`terions for formation of pharmaceutical salts list the counterions
`and their distribution in the respective data set only at a given
`point in time. Neither the distribution trends over time nor the
`causes for these have been analyzed to date.
`The present contribution examines the selection of counterions
`for the formation of salts by analyzing the Orange Book
`Database20 published by the U.S. Drug and Food Administration
`(FDA). The Orange Book lists all drug products approved in
`the U.S. Drug products approved after 1981 are listed including
`their date of approval. This enables an analysis of the changes
`in frequency of usage of the different counterions with time.
`Trends in salt selection over the past 25 years can thus be
`identified and the outcome of the overall analysis of the Orange
`Book compared to results based on other sources.
`
`Study Design
`
`Introduction
`Salt formation is a well-known technique to modify and
`optimize the physical chemical properties of an ionizable
`research or development compound. Properties such as solubil-
`ity, dissolution rate, hygroscopicity, stability, impurity profiles,
`and crystal habit can be influenced by using a variety of
`pharmaceutically acceptable counterions.1–8 Even polymorphism
`issues can be resolved in many cases by formation of salts. The
`crystal structure of a salt is usually completely different from
`the crystal structure of the conjugate base or acid and also differs
`from one salt to another. The modification of physical chemical
`properties, mainly solubility and dissolution rate, may also lead
`to changes in biological effects such as pharmacodynamics and
`pharmacokinetics,includingbioavailabilityandtoxicityprofile.1,9,10
`Owing to dramatic changes in the techniques applied in
`pharmaceutical discovery programs over the past 20 years, the
`physical chemical properties of development candidates have
`changed substantially.11 Drug design based on high-throughput
`screening has in general led to more lipophilic compounds
`exhibiting low aqueous solubility.
`There are many well-known formulation techniques to
`increase aqueous solubility,12–14 e.g., micronization, nanosizing,
`or complexation with cyclodextrins. The use of solid solutions
`and solid dispersions is another way to improve bioavailability
`for development candidates with low solubility. Nevertheless,
`formation of salts is almost
`the only chemical
`technique
`available to change aqueous solubility and dissolution rate
`without changing the API molecule. Further options for modify-
`ing these properties comprise the choice of the polymorphic
`form including solvates and formation of cocrystals. Although
`cocrystals in particular are an innovative way of designing APIs,
`this method is beyond the scope of this publication. An overview
`of this topic can be found in ref 15. Salt selection remains an
`important step at the interface between pharmaceutical research
`and development. A large number of publications covering
`
`The data were compiled from the FDA Orange Book
`Database as of the end of 2006. At this date, 21 187 drug
`products were listed, including 1356 chemically “well-defined”
`APIs. “Well defined” for the purpose of our analysis means
`that the API molecules are small chemical entities with a defined
`molar mass, typically below 1000 Da and that their chemical
`structure is completely known. Dosage forms containing
`multiple APIs, peptide hormones, biological APIs like antibod-
`ies, enzymes, extracts, and proteins, metal complexes, polymeric
`salt forms, inorganic APIs, and markers were excluded from
`our analysis. The APIs were classified into three categories:
`Category I consists of salts formed from basic molecules
`containing at least one atom suitable for protonation. Category
`II comprises salts formed from acidic species. Finally, category
`III is represented by APIs that are used as nonsalt forms. This
`* To whom correspondence should be addressed. Phone: +496151727634.
`class also includes zwitterions. Counterions are reported ac-
`Fax: +496151723073. E-mail: Christoph.Saal@merck.de.
`cording to their type of charge as cations and anions. The
`† Merck KGaA.
`stoichiometry of the salts is not discussed separately: for
`‡ Johann Wolfgang Goethe University.
`10.1021/jm701032y CCC: $37.00  2007 American Chemical Society
`Published on Web 12/01/2007
`
`Downloaded via REPRINTS DESK INC on October 28, 2020 at 23:21:16 (UTC).
`
`See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
`
`Liquidia's Exhibit 1024
`IPR2020-00770
`Page 1
`
`

`

`6666
`
`Journal of Medicinal Chemistry, 2007, Vol. 50, No. 26
`
`Paulekuhn et al.
`
`Table 1. Distribution of FDA Approved APIs among Categories I-III
`pre-
`1982
`(%)
`
`2002–2006
`(%)
`
`32.7
`
`14.6
`
`52.7
`
`overall
`(%)
`
`38.6
`
`38.4
`
`12.8
`
`13.6
`
`48.6
`
`48.0
`
`1997–2001
`1992–1996
`1987–1991
`1982–1986
`(%)
`(%)
`(%)
`(%)
`Category I: API Salts Formed of Basic Entities
`42.0
`40.2
`38.0
`40.3
`Category II: API Salts Formed of Acidic Entities
`10.1
`11.1
`13.3
`11.1
`Category III: Nonsalt APIs
`48.7
`48.7
`
`47.9
`
`48.6
`
`example, the occurrence of bromides includes bromides and
`dibromides. Furthermore, the APIs were arranged by year of
`approval to analyze how trends in the choice of salt forms have
`changed in recent decades. Prior to 1981, no date of approval
`is given in the Orange Book. Therefore, the drug products
`approved before 1982 are summarized under “pre-1982”. The
`period from 1982 to 2006 has been divided into five intervals,
`each comprising 5 years. After completion of the analysis of
`all chemically well-defined APIs, a separate assessment of the
`subset of APIs of oral (844 APIs) and injectable (482 APIs)
`dosage forms was made. Our analysis shows how the route of
`administration influences the choice of a specific salt form. This
`observation can be assigned to the different requirements of
`the two routes of administration. For example, for the two basic
`compounds biperiden and pentazocine, the chloride salts are
`used for oral dosage forms, whereas the lactate salts are used
`for injectable dosage forms.
`
`Results and Discussion
`Distribution of API Salts Formed of Basic and Acidic
`Molecules and APIs in Nonsalt Forms. The 1356 chemically
`well-defined APIs listed in the Orange Book comprise 659
`(48.6%) APIs in nonsalt forms, 523 (38.6%) salts formed from
`basic compounds, and 174 (12.8%) salts formed from acidic
`molecules. Thirty-eight different anions and 15 cations are used
`as counterions for the formation of salts. Thereof, 16 anions
`and 8 cations were only used once. During the past 25 years,
`25 anions and 7 cations have been used to form salts. The ratios
`of APIs obtained by salt formation of molecules exhibiting basic
`properties, API salts obtained from acidic species, and APIs in
`nonsalt forms have remained virtually constant. This is shown
`in Table 1. During 2002–2006, there has been some decrease
`in the percentage of APIs obtained as salts of basic compounds.
`This leads to a small increase in both of the other categories.
`Figure 1 shows the corresponding distribution of APIs among
`the three categories used in oral and injectable dosage forms.
`Together, oral and injectable formulations represent the majority
`of FDA-approved formulations. However, the requirements
`placed on an API for oral and injectable dosage forms are quite
`different. For oral dosage forms, a key prerequisite of the API
`is a certain minimum solubility in the pH range of the
`gastrointestinal
`tract. An adequate dissolution rate and a
`sufficient permeability are also important. If these requirements
`are not fulfilled, bioavailability will be insufficient to achieve
`the desired therapeutic effect. In the case of solutions for
`injection, considerations such as pH of the solution, osmolarity,
`and solubility in a small volume are important for efficient and
`pain-free administration. In many cases,
`this can lead to
`situations where a considerably higher solubility is required for
`injectables than for oral formulations.
`Distribution of Anionic Counterions Used To Form
`Pharmaceutical Salts. A summary of all anions used along
`with their distribution during different time periods is given in
`
`Figure 1. Classification and distribution of species in the Orange Book
`according to their type of charge and administration route.
`
`Table 2. Figure 2 displays the overall distribution of anions,
`whereas Figure 3 depicts the most recent period, 2002–2006.
`The anion encountered most frequently in FDA-approved
`pharmaceutical salts is the chloride ion. The fraction of chlorides
`increased from 52.9% (pre-1982) to 63.8% (1987–1991),
`remained almost constant at 63.3% over the next 5 years
`(1992–1996) and decreased significantly to 38.9% (2002–2006)
`over the past 10 years. The anion encountered with highest
`frequency after chloride is sulfate. However, it accounts for only
`7.5% of APIs formed from basic molecules. Its peak incidence
`was 12.0% during the period 1982–1986. Further acidic
`counterions frequently encountered include bromides, with a
`total incidence of 4.6%, as well as maleates and mesylates, both
`with incidences of 4.2%.
`There appears to be some tendency for “fashions” in anionic
`counterion selection, with certain counterions showing a notice-
`ably higher occurrence during one period compared to their
`overall usage. For example, nitrates represented 8.0% of anionic
`counterions during the 1982–1986 period. The average usage
`of nitrates is only 1.7%. Further examples include acetate with
`a maximum incidence of 12.7% during 1987–1991 and an
`overall usage of 3.3%. Tartrates exhibited a higher incidence
`of 6.7% in 1992–1996 than the average of 3.8%. Fumarates
`showed most frequent utilization during 1997–2001, contributing
`8.6% of FDA-approved salts formed of basic molecules during
`this period. They yielded an average fraction of 1.7%. For
`mesylates, the same is true with a peak occurrence of 13.8%
`during the same period and an average incidence of 4.2%. The
`number of anions used to form salts has varied during the past
`25 years between 11 and 15 per 5-year period. In total, there
`are only two anions with an average incidence of more than
`5% over the whole period. These are the chlorides and sulfates.
`Nevertheless, during the individual 5-year intervals, there are
`several anions reaching fractions of more than 5%. For example,
`in the pre-1982 period these are bromides and maleates. From
`1982 to 1986, acetates and nitrates are encountered in more
`than 5% of the APIs of category I. From 1987 to 1991, acetate
`and from 1992 to 1996 tartrate are the only anions other than
`chloride that were used to form more than 5% of the FDA-
`approved salts of basic molecules. After 1996, a broader variety
`of anions has reached an incidence of more than 5% usage.
`During 1997–2001 five anions exhibit an occurrence of more
`than 5%: bromides, chlorides, citrates, fumarates, and mesylates.
`From 2002 to 2006, seven different anions including bromides,
`chlorides, maleates, mesylates, phosphates, sulfates, and tartrates
`had an incidence of 5% or more. These figures indicate a strong,
`recent trend toward increased diversity of anions applied for
`the formation of salts in category I. The trend can be explained
`as a consequence of the changes in research techniques
`
`Liquidia's Exhibit 1024
`IPR2020-00770
`Page 2
`
`

`

`Trends in Salt Selection
`
`Journal of Medicinal Chemistry, 2007, Vol. 50, No. 26 6667
`
`Table 2. Distribution of Anions Used in APIs of Category I
`overall (%)
`pre-1982 (%)
`1982–1986 (%)
`3.3
`1.5
`8.0
`0.2
`0.8
`4.6
`0.2
`53.4
`0.2
`2.7
`0.2
`1.7
`0.2
`0.4
`0.2
`0.2
`1.0
`0.4
`1.3
`0.2
`0.2
`0.4
`4.2
`4.2
`0.4
`0.2
`0.4
`1.7
`0.2
`0.2
`0.2
`0.8
`2.7
`0.2
`1.2
`7.5
`0.2
`3.8
`0.4
`0.2
`
`acetate
`benzoate
`besylate
`bromide
`camphorsulfonate
`chloride
`chlortheophyllinate
`citrate
`ethandisulfonate
`fumarate
`gluceptate
`gluconate
`glucuronate
`hippurate
`iodide
`isethionate
`lactate
`lactobionate
`laurylsulfate
`malate
`maleate
`mesylate
`methylsulfate
`naphthoate
`napsylate
`nitrate
`octadecanoate
`oleate
`oxalate
`pamoate
`phosphate
`polygalacturonate
`succinate
`sulfate
`sulfosalicylate
`tartrate
`tosylate
`trifluoroacetate
`
`0.4
`5.2
`0.4
`52.9
`0.4
`2.6
`0.4
`0.4
`0.4
`0.7
`
`0.4
`1.5
`0.4
`1.5
`0.4
`0.4
`0.4
`5.5
`2.6
`0.7
`
`0.7
`0.7
`0.4
`
`1.1
`3.3
`0.4
`0.7
`9.6
`0.4
`3.7
`0.4
`
`2.0
`4.0
`
`52.0
`
`2.0
`
`2.0
`2.0
`4.0
`
`2.0
`2.0
`
`8.0
`
`12.0
`
`1987–1991 (%)
`12.7
`
`1992–1996 (%)
`
`1997–2001 (%)
`3.5
`1.7
`
`2002–2006 (%)
`2.8
`
`2.1
`
`63.8
`
`2.1
`
`2.1
`
`4.3
`
`2.1
`
`2.1
`
`2.1
`
`4.3
`
`2.1
`
`3.3
`1.7
`
`63.3
`
`3.3
`
`3.3
`
`1.7
`
`3.3
`1.7
`
`1.7
`
`1.7
`
`1.7
`
`3.3
`1.7
`
`6.7
`
`1.7
`
`5.2
`
`46.6
`
`5.2
`
`8.6
`
`3.5
`13.8
`
`1.7
`1.7
`
`1.7
`3.5
`
`3.5
`
`8.3
`
`38.9
`
`2.8
`
`2.8
`5.6
`8.3
`
`2.8
`
`2.8
`
`5.6
`
`2.8
`5.6
`
`8.3
`2.8
`
`number of salts
`
`523
`
`272
`
`50
`
`47
`
`60
`
`58
`
`36
`
`employed by the pharmaceutical industry. The extensive use
`of combinatorial chemistry and high-throughput screening in
`drug discovery has led to higher lipophilicity and commensurate
`lower solubility and dissolution rate of new drug candidates
`over the past 20 years. This in turn has necessitated a more
`intensive search for appropriate salts as a tool to improve
`physical chemical properties, a search typically conducted at
`the end of lead optimization or during exploratory development.
`
`Distribution of Cationic Counterions Used To Form
`Pharmaceutical Salts. All cationic counterions together with
`their respective incidences are listed in Table 3. Figure 4 shows
`the overall distribution of cations in salts formed from chemical
`entities exhibiting acidic properties. In Figure 5, the relative
`occurrence during the last period from 2002 to 2006 is depicted.
`Among the cations used to form API salts of acidic molecules,
`the sodium ion strongly dominates with an incidence of 75.3%
`over the entire period. From 1982 to 1991, the fraction of sodium
`salts was more than 90%. This decreased to 62.5% during the
`
`Figure 2. Overall distribution of anions used in APIs of category I in
`the Orange Book.
`
`Figure 3. Distribution of anions used in APIs of category I from 2002
`to 2006.
`
`Liquidia's Exhibit 1024
`IPR2020-00770
`Page 3
`
`

`

`6668 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 26
`
`Paulekuhn et al.
`
`Table 3. Distribution of Cations Used in APIs of Category I
`overall (%)
`pre-1982 (%)
`1982–1986 (%)
`0.6
`1.0
`6.9
`7.3
`0.6
`1.0
`0.6
`1.0
`0.6
`1.0
`0.6
`1.2
`2.9
`0.6
`6.3
`0.6
`0.6
`75.3
`1.7
`1.2
`
`benzathine
`calcium
`cholinate
`diethanolamine
`diethylamine
`lysine
`magnesium
`meglumine
`piperazine
`potassium
`procaine
`silver
`sodium
`tromethamine
`zinc
`
`5.2
`1.0
`6.3
`1.0
`1.0
`72.9
`
`1.0
`
`91.7
`
`8.3
`
`1987–1991 (%)
`
`1992–1996 (%)
`
`1997–2001 (%)
`
`2002–2006 (%)
`
`9.5
`
`14.3
`
`66.7
`9.5
`
`21
`
`18.8
`
`6.3
`6.3
`
`6.3
`
`62.5
`
`16
`
`6.3
`
`6.3
`
`87.5
`
`16
`
`92.3
`7.7
`
`13
`
`applied most frequently in APIs utilized in oral formulations is
`chloride. Its fraction increased from 55.8% (pre-1982) through
`65.4% (1982–1986) to 79.2% (1987–1991). After this period,
`there was a continuous decrease from 65.7% (1992–1996)
`through 45.0% (1997–2001) to 34.8% (2002–2006). Other
`important anions for oral delivery comprise sulfate with an
`incidence of 7.5%, maleate with 6.9%, and mesylate with 4.4%
`over the whole period. Mesylate salts exhibited a peak incidence
`of 15.0% during 1997–2001. Citrate salts were also frequently
`encountered during the same period, with 7.5% compared to
`an average fraction of 3.4% over the whole time period. The
`fifth anion according to frequency of usage ranking is bromide
`with an average value of 4.1% and a peak occurrence of 8.7%
`in 2002–2006.
`During each of the periods from 1982 to 1986 and 1987–1991,
`salts containing five different anions were approved in oral
`formulations. Between 1992 and 1996, 10 different anions were
`used in API salts in newly approved drug products intended
`for oral use. During the two last periods of 1997-2001 and
`2002–2006, 11 anions were applied per period. Thus, the overall
`trend toward a higher variety of acids and bases used for
`formation of salts is reflected in APIs for oral application.
`Distribution of Cationic Counterions Used in Oral
`Formulations. All cations encountered as counterions for
`formation of API salts used in products for oral delivery are
`summarized in Table 5. Sodium represents the most common
`cation of this category. Its average frequency of occurrence
`during the whole time period analyzed is 65.3%. It strongly
`fluctuates during the different 5-year time periods with a relative
`
`number of salts
`
`174
`
`96
`
`12
`
`2002–2006 period. The second most common cation is calcium
`with an average incidence of 6.9%. Its peak frequency of 18.8%
`was reached during 2002–2006. Another cation with frequent
`usage is potassium. On average, 6.3% of the FDA-approved
`drugs of category II are potassium salts. Potassium salts show
`their highest relative occurrence during 1992–1996, yielding
`14.3% of API salts obtained from acidic entities. Benzathine,
`cholinate, diethanolamine, diethylamine, meglumine, piperazine,
`procaine, and silver have not been used over the past 25 years.
`They were only used once each during the time frame before
`end of 1981. Lysine and magnesium were both introduced as
`counterions during the past 10 years.
`Only two basic counterions were utilized in each of the two
`5-year periods 1982–1986 (sodium, zinc) and 1987–1991
`(sodium, tromethamine). This number increased from three in
`the period 1997–2001 to five in the period 2002–2006. This
`analysis indicates that the trend toward a wider diversity of
`counterions observed for usage of anions is also occurring with
`cations.
`Salts Used in Oral Formulations. Of the 1356 chemically
`well-defined APIs listed in the Orange Book, 844 are used for
`oral delivery. A total of 449 (53.2%) of them are nonsalt forms,
`320 (37.9%) salts are formed from molecules exhibiting basic
`properties, and 75 (8.9%) are salts formed from entities with
`acidic behavior. A total of 30 different anions have been used,
`17 of them during the past 25 years. Only eight cations have
`been employed for formation of salts from acidic moieties, five
`of which were employed over the past 25 years. The analysis
`shows that 15 anions and 3 cations were only used once.
`Distribution of Anionic Counterions Used in Oral
`Formulations. Relative incidences of all anions used in FDA-
`approved oral formulations are presented in Table 4. The anion
`
`Figure 4. Overall distribution of cations used in APIs of category II
`in the Orange Book.
`
`Figure 5. Distribution of cations used in APIs of category II from
`2002 to 2006.
`
`Liquidia's Exhibit 1024
`IPR2020-00770
`Page 4
`
`

`

`Trends in Salt Selection
`
`Journal of Medicinal Chemistry, 2007, Vol. 50, No. 26 6669
`
`Table 4. Distribution of Anions for API Used in Oral Dosage Forms
`overall (%)
`pre-1982 (%)
`1982–1986 (%)
`0.9
`0.6
`7.7
`0.3
`0.6
`4.1
`56.6
`0.3
`3.4
`0.3
`1.6
`0.3
`0.3
`0.3
`0.3
`0.3
`0.3
`6.9
`4.4
`0.6
`0.6
`0.6
`0.3
`0.3
`0.9
`2.5
`0.3
`1.9
`7.5
`2.8
`0.3
`
`65.4
`
`3.9
`
`3.9
`
`19.2
`
`acetate
`benzoate
`besylate
`bromide
`chloride
`chlortheophyllinate
`citrate
`ethandisulfonate
`fumarate
`gluconate
`hippurate
`iodide
`lactate
`laurylsulfate
`malate
`maleate
`mesylate
`methylsulfate
`napsylate
`nitrate
`octadecanoate
`oxalate
`pamoate
`phosphate
`polygalacturonate
`succinate
`sulfate
`tartrate
`tosylate
`
`0.6
`5.2
`55.8
`0.6
`4.1
`0.6
`0.6
`0.6
`0.6
`0.6
`0.6
`0.6
`
`8.7
`1.7
`1.2
`1.2
`
`0.6
`
`1.7
`2.9
`0.6
`1.2
`7.6
`1.7
`
`number of salts
`
`320
`
`172
`
`26
`
`Table 5. Distribution of Cations for API Used in Oral Dosage Forms
`
`1987–1991 (%)
`
`1992–1996 (%)
`
`1997–2001 (%)
`
`2001–2006 (%)
`
`79.2
`
`4.2
`
`8.3
`
`4.2
`4.2
`
`24
`
`2.9
`
`65.7
`
`2.9
`
`2.9
`
`5.7
`2.9
`
`2.9
`
`5.7
`2.9
`5.7
`
`35
`
`2.5
`
`5.0
`45.0
`
`7.5
`
`5.0
`
`5.0
`15.0
`
`2.5
`
`2.5
`5.0
`5.0
`
`40
`
`8.7
`34.8
`
`4.4
`8.7
`8.7
`
`4.4
`
`8.7
`
`4.4
`8.7
`4.4
`4.4
`
`23
`
`benzathine
`calcium
`cholinate
`magnesium
`piperazine
`potassium
`sodium
`tromethamine
`
`overall (%)
`1.3
`12.0
`1.3
`2.7
`1.3
`13.3
`65.3
`2.7
`
`number of salts
`
`75
`
`pre-1982 (%)
`2.3
`11.4
`2.3
`
`2.3
`13.6
`68.2
`
`44
`
`1982–1986 (%)
`
`1987–1991 (%)
`
`1992–1996 (%)
`
`1997–2001 (%)
`
`2002–2006 (%)
`
`11.1
`
`33.3
`44.4
`11.1
`
`9
`
`11.1
`
`88.9
`
`9
`
`50.0
`
`16.7
`
`16.7
`16.7
`
`6
`
`100.0
`
`1
`
`83.3
`16.7
`
`6
`
`fraction of at least 68.2% until 1991. This value decreased to
`44.4% during 1992–1996. During the following period,
`1997–2001, there was an increase to 88.9% followed by a huge
`drop to just 16.7% during 2002–2006. The strong fluctuations
`are caused by the small absolute numbers of approved drug
`products containing salts formed from acidic entities. There were
`a maximum of nine drugs approved in this category for oral
`usage during each of the 5-year periods. The second common
`cation is potassium with an average fraction of 13.3% over the
`whole period and a peak of 33.3% in 1992–1996. The third
`important cation for oral dosage forms, which accounted for a
`total frequency of 12.0% and a peak of 50.0% during the last
`period from 2002 to 2006, is calcium. Thus, calcium and
`potassium have changed positions in usage ranking for oral
`dosage forms in recent times.
`A good example of how the counterion affects the physical
`chemical properties of an API in oral formulations is diclofenac
`and its salts. There are both sodium and potassium salts of
`diclofenac applied in drug products for oral delivery. The free
`acid is not used in FDA-approved drug products. Only the
`diclofenac sodium salt is utilized for extended and delayed
`release tablet dosage forms. In contrast, the diclofenac potassium
`salt is used for immediate release tablets. This suggests that
`
`the different salt forms may influence dissolution rates. Fini et
`al.21 have discussed the difference in dissolution behavior
`between these salt forms.
`Salts Used in Injectable Formulations. The 482 APIs used
`for injectable formulations consist of 171 (35.5%) nonsalt forms,
`208 (43.2%) API salts of basic molecules, and 103 (21.4%)
`salts of acidic entities, whereas in APIs utilized in oral
`formulations about half of the APIs were used as nonsalt forms;
`in injectable formulations only about one-third were employed
`as noncharged forms. This shows that formation of salts is even
`more important for injectable dosage forms than for oral
`formulations. The more frequent usage of salt forms in injectable
`formulations can be explained by the need for even higher
`solubility compared to oral formulations. An oral dosage form
`needs to completely dissolve in 250 mL of aqueous media in
`the physiological relevant pH range of 1–8 to be classified as
`highly soluble with reference to the Biopharmaceutical Clas-
`sification System.22 Typically, the preferred injectable dosage
`form comprises a volume of a few milliliters. If the solubility
`of the API is too low for this application, an infusion formulation
`becomes necessary. In many cases, there is a difference of at
`least one order of magnitude with respect to the solubility
`required for the formulation of an API as an injectable versus
`
`Liquidia's Exhibit 1024
`IPR2020-00770
`Page 5
`
`

`

`Table 6. Distribution of Anions for API Used in Injectable Dosage Forms
`overall (%)
`pre-1982 (%)
`1982–1986 (%)
`5.8
`2.3
`5.0
`1.4
`0.8
`5.0
`4.3
`3.9
`5.0
`0.5
`0.8
`53.4
`54.3
`0.5
`0.8
`2.4
`1.6
`0.5
`0.8
`0.5
`0.5
`0.5
`0.5
`1.0
`1.0
`2.9
`0.5
`0.5
`1.4
`3.9
`0.5
`0.5
`0.5
`3.4
`0.5
`8.2
`3.9
`0.5
`0.5
`
`60.0
`
`5.0
`
`5.0
`5.0
`
`10.0
`
`0.8
`0.8
`-
`1.6
`0.8
`3.1
`0.8
`0.8
`2.3
`3.1
`0.8
`
`3.9
`
`10.9
`4.7
`0.8
`
`acetate
`besylate
`bromide
`camphorsulfonate
`chloride
`chlortheophyllinate
`citrate
`ethandisulfonate
`fumarate
`gluceptate
`gluconate
`glucuronate
`iodide
`isethionate
`lactate
`lactobionate
`malate
`maleate
`mesylate
`nitrate
`oleate
`pamoate
`phosphate
`succinate
`sulfate
`tartrate
`tosylate
`trifluoracetate
`
`5.3
`
`42.1
`
`5.3
`
`5.3
`
`5.3
`
`5.3
`5.3
`
`5.0
`5.0
`
`55.0
`
`5.0
`
`5.0
`
`5.0
`
`5.0
`5.0
`
`5.0
`
`5.0
`
`20
`
`7.1
`
`50.0
`
`21.4
`
`7.1
`
`50.0
`
`16.7
`
`16.7
`
`14
`
`6
`
`6670
`
`Journal of Medicinal Chemistry, 2007, Vol. 50, No. 26
`
`Paulekuhn et al.
`
`1987–1991 (%)
`26.3
`
`1992–1996 (%)
`
`1997–2001 (%)
`14.3
`
`2002–2006 (%)
`16.7
`
`number of salts
`
`208
`
`129
`
`20
`
`19
`
`Table 7. Distribution of Cations for API Used in Injectable Dosage Forms
`overall (%)
`pre-1982 (%)
`1982–1986 (%)
`1.0
`1.6
`2.9
`4.8
`1.0
`1.6
`1.0
`1.6
`1.0
`4.9
`1.0
`1.0
`85.4
`1.0
`
`7.9
`1.6
`1.6
`79.4
`
`100.0
`
`benzathine
`calcium
`diethanolamin
`diethylamin
`lysine
`meglumine
`potassium
`procaine
`sodium
`tromethamine
`
`number of salts
`
`103
`
`63
`
`9
`
`an oral dosage form, with higher solubility generally required
`for APIs used in injectable dosage forms. The increased
`percentage of APIs employed as salt forms in injectable dosage
`forms shows that formation of salts is a practical way to achieve
`this objective. A total of 28 different anions and 10 different
`cations were used as counterions for formation of salts utilized
`in FDA-approved injectable formulations. Seventeen anions and
`only three cations were used over the past 25 years.
`Distribution of Anionic Counterions Used in Injectable
`Formulations. A summary of the frequency of occurrence of
`all anions used for the formation of salts of basic molecules in
`injectable formulations is presented in Table 6. As for oral
`dosage forms, the most important anion is chloride with an
`average fraction of 53.4%. This incidence has remained quite
`stable, exhibiting a minimum of 42.1% and a maximum of
`60.0%. During the last two periods (1997–2001 and 2002–2006)
`the fraction was 50.0% each. The second widely used anion is
`sulfate with a total fraction of 8.2%. However, after 1991 no
`further sulfate salts have been approved for injectable dosage
`forms. The third anion in frequency of occurrence ranking is
`acetate with an average fraction of 5.8% and a peak value of
`26.3% during the 1987–1991 period. During the following
`
`1987––1991 (%)
`
`1992–1996 (%)
`
`1997–2001 (%)
`
`2002–2006 (%)
`
`88.9
`11.1
`
`9
`
`100.0
`
`100.0
`
`8
`
`7
`
`14.3
`
`85.7
`
`7
`
`period, from 1992 to 1996, there were no further FDA-approved
`acetate salts. On the other hand, during the last two periods
`1997–2001 and 2002–2006 the relative fraction of acetates
`increased to 14.3% and 16.7%. Frequent usage of mesylates
`over the past 10 years, with a relative frequency of occurrence
`of 21.4% during the 1997–2001 period and 16.7% during the
`2002–2006 period, is apparent from Table 6. This is in strong
`contrast to the period from 1982 to 1996 in which no mesylate
`salts were approved for injectable dosage forms. In contrast to
`API salts containing anionic counterions intended for oral
`formulations, a trend toward a broader variety of anions cannot
`be observed for injectable formulations.
`Distribution of Cationic Counterions Used in Injectable
`Formulations. In category II, 38 of the 40 APIs used in
`injectable formulations and approved over the past 25 years are
`sodium salts. Beyond the sodium salts,
`there is only one
`tromethamine salt approved in 1989 and one lysine salt approved
`in 2006. A summary together with the 63 salt forms approved
`before 1982 is given in Table 7.
`Comparison with Analysis of Data from the Cam-
`bridge Structural Database. Haynes, Jones, and Motherwell
`searched the Cambridge Structural Database (CSD) for the
`
`Liquidia's Exhibit 1024
`IPR2020-00770
`Page 6
`
`

`

`Trends in Salt Selection
`
`Journal of Medicinal Chemistry, 2007, Vol. 50, No. 26 6671
`
`occurrence of salts with pharmaceutically acceptable counter-
`ions.23 It is mentioned that the CSD is a database that is not
`limited to pharmaceuticals. Rather, it contains many substances
`used in other industries, such as pigments. The analysis of
`Haynes et al. was published in 2005, covering a time span of
`more than 80 years. Haynes et al. received 6021 hits for anions
`and 587 hits for cations. A hit represents one structure of an
`organic salt found in the CSD. Because of the fact that the CSD
`is not a database exclusively comprising APIs, it is difficult to
`obtain pharmaceutically relevant trends in salt selection from
`this database.
`Haynes et al. searched the CSD for salt forms containing
`pharmaceutically acceptable counterions. For this search they
`used 69 different anions and 21 different cations. However, since
`the authors faced difficulties in determining charges and the
`bonding type of metal atoms, they were unable to differentiate
`appropriately between ionic and covalent compounds. This
`problem forced the authors to omit all compounds containing
`metal atoms. Because metal cations are the most frequently used
`cationic counterions in the Orange Book, a comparison of the
`data between the Orange Book and CSD for cations is not
`meaningful.
`As a consequence, only the results for anionic counterions
`are compared with the Orange Book data. The comparison of
`the relative occurrence of anions used as counterions for the
`formation of salts shows large differences between the CSD
`and the Orange Book analysis. As one example, bromides used
`for formation of salts account for a much higher share in the
`CSD (23.3%) than in the Orange Book (4.6%). In contrast to
`this observation, the results for chlorides agree quite well: 47.7%
`in the CSD and 53.4% in the Orange Book. The maleate,
`mesylate, and sulfate fractions in the CSD are distinctly lower
`than in the Orange Book: 1.3% (CSD) versus 4.2% (Orange
`Book) for maleates, 1.1% (CSD) versus 4.2% (Orange Book)
`for mesylates, and 2.7% (CSD) versus 7.5% (Orange Book) for
`sulfates.
`The ratio of salts formed with anionic counterions to salts
`formed with cationic counterions in the CSD analysis is about
`10 to 1. The respective ratio obtained from the Orange Book is
`roughly 3 to 1. This reflects the large fraction of compounds
`left out by neglecting substances containing metal cations in
`the CSD analysis. Nonsalt forms of API were not considered
`in the CSD analysis.
`The CSD analysis for cationic counterions loses pharmaceuti-
`cal relevance by using a database that
`includes non-API
`substances and leaves out metal cations as counterions. Surpris-
`ingly, the analysis for anionic counterions gives the right order
`of magnitude for most anions. Nevertheless, examples such as
`the bromide salts show that the CSD results are not sufficiently
`reliable. In conclusion, analysis of a very general database like
`the CSD cannot be expected to and does not yield results
`relevant in a pharmaceutical environment.
`Comparison with Analysis of Data from Martindale.
`Berge, Bighley, and Monkhouse published a review article about
`pharmaceutical salts in 1977.1 In this article, the distribution of
`counterions at that time was presented. Their list was based on
`Martindale’s “The Extra Pharmacopoeia”, 26th edition, from
`1974. The authors listed 80 different anions and 21 different
`cations used as counterions for formation of pharmaceutical
`salts. At that time, 53 anions and 14 cations were classified as
`FDA-approved. The distribution of counterions obtained in this
`analysis is comparable to the average values from the Orange
`Book compilation obtained 30 years later. This can be derived
`from the data summarized in Table 8. The good agreement is
`
`Table 8. Comparison of Orange Book (2006) Data with Data from
`Berge, Monkhouse, and Bighley (1993 and 1974)
`
`counterion
`bromide
`chloride
`maleate
`mesylate
`sulfate
`calcium
`potassium
`sodium
`
`Martindale,
`1974 (%)
`7.6
`47.7
`3.0
`2.0
`7.8
`10.5
`10.8
`62.0
`
`Martindale,
`1993 (%)
`5.7
`48.9
`3.1
`3.2
`6.1
`12.2
`9.8
`57.7
`
`Orange Book,
`2006 (%)
`4.6
`53.4
`4.2
`4.2
`7.5
`6.9
`6.3
`75.3
`
`not surprising because the trend toward a broader variety of
`counterions first started to have a notable impact on distributions
`around the mid-1990s. Because of the large number of APIs
`approved before that point in time, the average distribution is
`still do