320 ANALYTICAL PROCEEDINGS, SEPTEMBER 1086, VOL 23 Short Papers in Pharmaceutical Analysis The following is a summary of one of the papers presented at a Meeting of the Joint Pharmaceutical Analysis Group held on October loth, 1985, at The Pharmaceutical Society of Great Britain, London SEI. Summaries of eight other papers presented at the Meeting appeared in the July issue, p. 254. Stability of Pralidoxime Mesylate Injections D. G. Holcombe The Laboratory of the Government Chemist, Cornwall House, Waterloo Road, London SE? 8XY The value of certain oximes as antidotes against organophos- phorus nerve gases and pesticides, both of which act by the inhibition of cholinesterase, is well established. The most widely used oxime is pralidoxime (2-hydroxyiminomethyl-l- methylpyridinium), either as the chloride, abbreviated to 2PAM (US Pharmacopeia XX), or as the mesylate (methyl- sulphonate), abbreviated to P2S (Nordic Pharmacopoeia). The normal strength used is 1 g in 5 ml for intramuscular injections, and stocks of P2S are kept by emergency holding centres throughout the country. The Laboratory of the Government Chemist is responsible for monitoring the stabil- ity of these stocks on behalf of the Department of Health and Social Security (DHSS).In the early 1960s, there was extensive investigation of the possible routes of degradation of pralidoxime in solution. It was thought that it could undergo acid hydrolysis to form 2-formyl- 1-methylpyridinium (“aldehyde”) plus hydroxyl- amine, or alkaline hydrolysis to 2-cyano-1-methylpyridinium (“nitrile”), and thence either to 2-methylpyridone (“pyri- done”) or via 2-carbamoyl-1-methylpyridinium (“amide”) to 2-carboxy- 1-methylpyridinium (“acid”). 1 These studies relied primarily on TLC and UV techniques.Although the existence of isomeric oximes was discussed in 1957,* subsequent IR studies suggested that pralidoxime only contained one isomer, thought to be the syn form. In this laboratory, P2S has been assayed by the UV method of the US Pharmacopeia,3 supplemented by TLC for degradation, IR for identification and a limit test for cyanide. A stability indicating HPLC system was developed, which separated P2S from four degradation products.4 The HPLC system reported by Uttley,s incorporating amine modification of the mobile phase, improved the separation and resolved a new peak, which was proposed as an isomeric form of P2S-the anti form. We used a related system to investigate this new peak and its significance to the stability of P2S.Experimental The following equipment and conditions were used: HPLC pump, Waters 6000A; loop, Rheodyne, 20 pl; column, 250 X 4.6 mm, 5 pm Spherisorb ODs; mobile phase, methanol 25%, aqueous 75% (diethylamine phosphate 0.05 M , sodium lauryl sulphate 0.01 M, pH 3); detector, Beckman 165 dual channel UV detector, at 265 and 293 nm (with facility to scan peaks “on the fly”). Results and Discussion The amine-based HPLC system, described under Experimen- tal, resolved the extra peak that eluted between the amide and syn P2S (Fig. 1). Using the scanning facility on the UV 265 n x 32 ”! 293 nm x 8 b ? cn I 05 1 n 0 e Fig.1. Separation of P2S from degradation products using an amine-modified HYLC system. 1, P2S “syn”; l a , P2S ”arzti”; 2, aldehyde (not resolved); 3, pyridone; 4, acid; 5 , amide; 6, nitrile detector, P2S showed a maximum absorbance at 293 nm for the syn isomer and at 265 nm for the anti isomer (Fig. 2). The UV spectrum of the anti isomer resembles that of the “aldehyde,” but measurement of the level of hydroxylamine rules out the extra peak being the aldehyde. It is unusual that two stereoisomers should show such different spectra. The most probable explanation is that with the syn isomer, the wavelength of the basic chromophore is lengthened to 293 nm by conjugation with the side-chain. In the anti isomer, steric interaction between the N-methyl and the oxime hydroxy group forces the aldoxime out of the plane of the ring and reduces the wavelength of the chromophore to 265 nm.The conversion of syn P2S to the anti form was enhanced by UV irradiation in 10 mg ml-1 concentrati0n.j A maximum yield of 55% anti P2S was obtained after 24 h. After this time, degradation of the anti P2S to “amide” and “nitrile” seemed to predominate. Using dilute standards of P2S (0.1 mg ml-l. pH 5 . 5 ) , it was found that conversion to the anti form could be achieved in sunlight ( e . g . , daylight). In a timed study of the degradation of these standards, the level of anti F2S was 9.9% after 4 h, reaching a maximum of 54% after 7 d. At this time only 9.3% of the original syn P2S appeared to remain.“Amide” and “nitrile” were observed from 4 d onwards and traces of “pyridone” and “acid” were present after 7 d. After 7 d the “aldehyde” level was 0.18% (from the measurement of hydroxylamine). Acidification of 0.1 mg ml-1 solutions to pH 3.2 slowed down the rate of photolytic conversion to the anti form to one third of its original value. Protection of 0.1 mgANALYTICAL PROCEEDINGS, SEPTEMBER 1986, VOL 23 32 1 ml-1 solutions, pH 5.5, from daylight appeared to prevent conversion to the anti form over an 8 d period even when not refrigerated. Thus, the rate of conversion appears to depend on sunlight, dilution factor and pH. 2 00 250 300 350 400 i. nrn Fig. 2. Normalised UV spectra of P2S isomers. A, syn; B. anfi A pure standard of anti P2S was not obtainable, but approximate El values were calculated; 176 at 265 nm (169 for syn) and 154 at 293 nm (532 for syn).In sunlight, the photolytic mechanism of conversion could be the excitation of electrons in the C=N bond into the JT;* anti-bonding orbital. If this state can exist long enough for rotation to occur about the C-N bond, then conversion can proceed from the syn to the anti form. The increased isomeric conversion in more dilute solutions is attributable to reduced competition among the P2S molecules for the excitation energy. It was previously thought that alkaline hydrolysis to the amide occurred via the “nitrile.”’ It is likely, however, that syn P2S first converts to “anti” P2S and may then follow one of two paths (Fig. 3): ( a ) it may undergo a Beckman rearrangement to an “enol,” which then forms the “amide.” or ( b ) as anti P2S has the correct geometry to eliminate water across the C=N, forms the “nitrile.” Thus, “amide” and “nitrile” may be formed by independent routes. Beckmann rearrangement of syn P2S is unlikely since it necessitates migration of the pyridinyl ring. The long term stability of P2S has been extensively investi- gated.4 In a recent study on expired ampoules (20% rnlV). it was found that the pH ranged from 3.2 to 3.5 and isomeric conversion did not exceed 3.5% of total P2S; degradation was not significantly increased over a 7 d period when ampoules were deliberately exposed to sunlight. All samples showed traces of “amide,” but *‘nitrile” in only some. Trans Elimination of water - CH3 Nitrile Anti P2S Beckrnann I rearrangement I t CH3 I t NH1 Enol Arnide Fig, 3.Formation o f amide and nitrile from anti P2S Proton NMR spectroscopy of an expired sample was inconclusive because the syn signals masked possible isomeric shifts. However, decoupled 13C studies showed four isomeric shifts out of a possible six, inferring the presence of an isomeric oxime. No “aldehydic” carbons were observed. Conclusion In solution, P2S appears to exist in two stereoisomeric forms. The syn isomer normally predominates, but, under certain conditions, converts to the anti form, which is apparently the precursor to the degradation pathways. Current formulations have been seen to be acceptably stable with respect to isomeric conversion and degradation. The author thanks the Scientific and Technical Branch of the DHSS €or financing this project. References 1. 2. 3. 4. 5 . Brown, N. D., Strickler, M. P.. Sleeman, H. K . , and Doctor. B. P., J. Chromatogr., 1981. 212, 361. Ginsburg, S . , and Wilson. I. B . J . Am. Chem. Soc.. 1957. 79. 481. “United States Pharmacopeia, XXth Revision, National Formu- lary, XVth Edition,” Mack, Easton, PA. 1980, pp. 648-649. May, E. M., and Pearse, J. E.. Anal. Proc., 1983. 20, 179. Utley, D., J. Chromatogr.. 1983, 265, 311.