The Importance of a Salt Forming Agent’s Molecular Weight in choosing a Pharmaceutical Salt

In Stahl and Wermuth’s book, Pharmaceutical Salts: Properties, Selection and Use there is a further piece of advice beyond what Kilomentor has already written about concerning salt selection. Unlike the other advice it is provided by implication only and needs to be simply stated. 

On pg. 181 of the book, the selection of an appropriate pharmaceutical salt for the candidate drug called RPR200765 is presented.  The following details of that problem are provided. RPR200765 was a candidate drug substance to be used to treat rheumatoid arthritis. The drug would have had to be taken regularly for the rest of a patient’s life.  It is a crystalline, weak base with a substituted pyridine ring system, a pKa of 5.3 and log P of 2.5.  The anticipated pharmaceutically effective dose was expected to fall between 100-125 mg.  One can calculate that the molecular weight of RPR200765 by itself was 488.48. The actual API material is identified in Bioorganic & Medicinal Chemistry Letters (200), 11(5) 693-696.

Four potential salts were identified in the example: mesylate, camphorsulfonate, hydrochloride, and hydrobromide.  What was particularly instructive is the comment concerning the camphorsulfonate. The authors wrote that the only disadvantage of the camphorsulfonate, when compared to the mesylate (the first choice), was the increased molecular weight due to the larger counter ion. It was considered that this could create problems with experimental capsule or tablet formation later in development.

Camphorsulfonic acid has a molecular weight of 232. The molecular weight of the mono camphorsulfonate salt of RPR200765 would have been 720.48. Giving a dose of 100-125 mg on the free base basis (0.256 mmoles), as camphorsulfonate salt, would amount to giving a dose of 184 mg of this pharmaceutical salt.  Delivering a dose of 184 mg of API, it is said, was anticipated to be problematic.  From this, it is possible to generalize that the practical limit to the weight of API that can be confidently handled is about 184 milligrams in the highest strength.  This seriously restricts the choices of pharmaceutical salts for medicines particularly where the neutral active has a low molecular weight because this means there will be more moles in the dose and so more moles of salt former.

To take a current example, the highest prescribed dose of the cancer drug imatinib is 400 mg as the free base. The molecular weight of the free base is 493. If we imagine the salt with an acid of molecular weight 232(camphorsulfonate) the weight of active API would be 588 mg. This is already more than 3 times what this teaching advises one can be comfortable with for achieving a successful formulation. It is obvious that the acid used to form the pharmaceutical salt for imatinib is going to have to have a low molecular weight.  Pharmaceutically acceptable acids with a molecular weight below 100 are only: acetic, carbonic, formic, glycolic, hydrobromic, hydrochloric, isobutyric, lactic, methanesulfonic, nitric, oxalic, phosphoric, sulphuric, and thiocyanic.  Of these, the only ones without other concerns are hydrochloric, methanesulfonic, phosphoric, and sulphuric.  Suddenly salt selection becomes a lot easier! In the case of imatinib, the methanesulfonate was chosen as the drug substance!

Put  conversely this means that besides camphorsulfonic acid also galactaric, glucoheptanoic, lactobionic, 2-naphthalenesulfonic, 1,5-naphthalenesulfonic, oleic, palmitic, pamoic, sebacic, stearic, and tannic acids need not be initially considered for salt formation with candidate bases. Five of these are from the group of 30 called Class 1 acids, the most preferred acids based on safety considerations.

Similarly benethamine, benzathine and hydrabamine are distinctly less preferred for salt formation with acid candidates based on their molecular weights.