Use of thin-layer chromatography to detect counterfeit sulfadoxine/pyrimethamine tablets with the wrong active ingredient in Malawi

Out of the 13 districts of southern Malawi, four were randomly selected. From each of these districts, a list of government health centers was obtained. From each of three districts, two health centres were selected randomly, and samples were collected from these health centres as well as from the respective district hospital. The fourth district comprised one of Malawi’s larger cities. From this district two urban and two rural health centres were randomly selected, and samples were collected from there, from the district health office and from the central hospital. If church-affiliated health facilities, private pharmacy shops, drug stores, or illegal street vendors could be identified nearby the selected government health facilities, then drugs samples were also collected from there. Samples were collected by members of the Pharmacy Department, University of Malawi. A mystery shopper approach was used for the illegal street vendors, and an overt approach for the other facilities. The mystery shopper stated that he had been asked by friends in his village to buy this medicine for them. Samples of 150 tablets were collected if available, otherwise smaller numbers. If medicines with generic and brand name were available, the brand name medicine was sampled. If several brand name medicines were available, the most expensive brand name medicine was sampled. In most sites, however, only a single type of sulfadoxine/pyrimethamine tablets was available. Samples were transported to the Pharmacy Department, College of Medicine, Blantyre, within 48 h, and stored below 25 °C until analysis.

The external packaging, primary packaging and (if available) package leaflets were inspected, including batch number and expiry dates. The tablets were visually inspected, especially for undamaged, unaltered surfaces and colour uniformity. Disintegration testing for instant-release oral dosage forms was carried out according to the manual of the GPHF Minilab [9]; in short, six tablets were kept in water at 37 °C under occasional shaking or stirring, and complete disintegration within 30 min was confirmed. For uniformity of mass of dosage unit, the exact weight of 20 tablets was determined; acceptable deviations were: up to ±5 % in at least 18 tablets, and up to ±10 % in no more than two tablets.

TLC testing was done according to the procedure given by the manual of the GPHF Minilab for sulfadoxine (including SP formulations) [9]. From each sample, three tablets were analysed individually. In short, each tablet was crushed to a fine powder and extracted with 20 ml methanol by vigorous shaking for 3 min. After sedimentation of undissolved residues, 1 ml of the supernatant was removed and diluted with 3 ml methanol. Using a microcapillary, 2 µl of this solution were applied to a TLC plate (Merck silica gel 60 F254, 0.2 mm thickness, 5 × 10 cm). Authentic standard solutions of sulfadoxine/pyrimethamine with known concentrations were applied as comparison. The plate was developed in a solvent system of ethyl acetate:methanol 15:5 for approximately 15 min. After drying off the residual solvent, the active pharmaceutical ingredients were visualized first under UV light (254 nm), and subsequently by iodine vapour. The results were documented using an inexpensive digital camera (Canon PowerShot SX600 HS).

For comparison to authentic paracetamol and co-trimoxazole samples, the solvent systems given by the manual of the GPHF Minilab for analysis of paracetamol[9], pyrimethamine [10] and co-trimoxazole [9] were used, i.e. acetone:toluene:acetic acid 10:10:0.5 (for the experiment depicted in Fig. 2b); ethyl acetate:methanol:acetone:conc.aqueous ammonia 12:6:2:0.5 (Fig. 2c); ethyl acetate:methanol 15:5 (Fig. 2d).

Following the methods specified in USP38-NF33, identification of the active ingredients by TLC and HPLC, HPLC analysis (=assay) for the content of sulfadoxine and pyrimethamine, analysis for uniformity of dosage units with respect to the content of the active ingredients, and analysis of their dissolution was carried out in the WHO-prequalified drug quality control laboratory of the Mission for Essential Drugs and Supplies, Nairobi, Kenya. HPLC analysis for sulfadoxine and pyrimethamine was carried out using a Gemini 5 µm C6-Phenyl 110Å HPLC column 250 × 4.6 mm (Phenomenex, USA) and an isocratic solvent system of 0.1 % aqueous phosphoric acid:acetonitrile 83:17, flow rate 1.2 ml/min. The wavelength for UV detection was 230 nm. For the identification of paracetamol and co-trimoxazole, the respective methods of USP38-NF33 for those drugs were used.

This study was approved by the College of Medicine Research and Ethics Committee, University of Malawi.

Twenty eight samples of sulfadoxine 500 mg/pyrimethamine 25 mg tablets were collected in four districts in southern Malawi. 15 were collected from government health facilities, seven from church-affiliated health facilities, four from private pharmacies and drug stores, and two from illegal street vendors. 21 of the samples were found to be distributed under the generic name “sulfadoxine/pyrimethamine”, and seven under brand names. According to the information on the packaging, the samples had been produced by six different manufacturers, with 18 samples produced in in India (by two different manufactures), three in China, three in Tanzania, two in Cyprus and two in Malawi. Comparison with the records of the Pharmacy, Medicines and Poisons Board, i.e. the national drug regulatory agency of Malawi, showed that the SP tablets from five of the manufacturers were registered in Malawi, but the tablets from one of the manufactures were not. The non-registered type represented the most common SP preparation collected in government and church-affiliated health facilities, accounting for 17 of the 28 samples.

Only a single sample clearly failed visual inspection. It was purchased from an illegal street vendor and was sold in an opened, apparently genuine cardboard box labelled “Novidar (SP)”, a brand name of SP manufactured and sold by the Malawian manufacturer Pharmanova Ltd. The cardboard box contained paper strip packs labelled “Novidar (SP)” which, however, were found to be of two different kinds (Fig. 1). One (hereafter called type N) was stamped with the same batch number and expiry date as given on the outer packaging (i.e. the cardboard box). The other one was stamped with two dates (“27/04/2010” and “20/11/2015”), different from those given on the outer package. These paper strip packs were made from a thinner type of paper than those of type N. A part of the tablets in these strip packs had apparently adsorbed moisture. Upon opening of the strip packs, some tablets were found to stick to the paper, and to break easily.

Fifteen paper strips of this kind were contained in this sample. Although they were all uniform in their appearance and stamp, they were found to contain two different kinds of tablets (Fig. 1). One (hereafter called type X) did not carry an imprint on its front side. The other one (hereafter called type Y) was imprinted with the letters “UCL”. In contrast, the tablets with the paper strips of type N were imprinted with “Novidar SP” on the front side, which is consistent with the genuine product of Pharmanova Ltd.

The tablets of types N, X and Y were subjected to thin-layer chromatographic (TLC) analysis according to the procedure given in the manual of the GPHF Minilab for sulfadoxine/pyrimethamine tablets (see Methods section). An authentic standard of sulfadoxine 500 mg/pyrimethamine 25 mg was used for comparison. Detection was carried out first with UV light (254 nm), then with iodine staining. The result is shown in Fig. 2a. The complete analytical procedure was repeated again, starting from different tablets. The results were identical to those shown in Fig. 2a.

Both in the first and the second analysis, the tablets of type N showed spots identical in Rf value and intensity to those of sulfadoxine and pyrimethamine in the authentic standard. This strongly indicates that this product contains the declared active ingredients in the declared amounts, and most likely represents the original product “Novidar (SP)” of the Malawian manufacturer Pharmanova Ltd. Type X showed no spots identical in Rf value to those of sulfadoxine and pyrimethamine. This proves the absence of relevant quantities of both active principles in this product. Instead, TLC analysis showed another compound, giving a spot of an Rf value slightly lower than sulfadoxine. The different Rf value, and the different response to iodine staining (Fig. 2a), prove that this compound is different from sulfadoxine.

Type Y showed no spot identical in Rf value to that of pyrimethamine from the authentic standard. This proves the absence of relevant quantities of pyrimethamine in this product. However, TLC analysis did show a spot similar in Rf value and intensity to that of sulfadoxine, indicating the presence either of sulfadoxine or of a compound with similar chromatographic behaviour. Furthermore, TLC analysis showed a spot of a further compound, with an Rf value clearly lower than pyrimethamine.

The imprint “UCL” is used by the Kenyan pharmaceutical manufacturer Universal Corporations Ltd. Therefore, tablets of type Y were compared the with drugs from this manufacturer. And indeed, Sulfran™ tablets (co-trimoxazole 480 mg) marketed by UCL in Malawi were found to be perfectly identical in shape, size and imprint to the tablets of type Y. It was furthermore speculated that type X may represent paracetamol tablets.

To test these hypotheses, the tablets were further analyesed in comparison to authentic paracetamol 500 mg tablets and to co-trimoxazole 480 mg tablets using the TLC solvent systems given in the manual of the GPHF Minilab for paracetamol (Fig. 2b) for pyrimethamine (Fig. 2c) and for sulfamethoxazole and co-trimoxazole (Fig. 2d). In all three solvent systems, type X showed identical results as paracetamol 500 mg tablets, and type Y showed identical results as co-trimoxazole 480 mg. Both type X and type Y proved to be clearly different from their declared content, i.e. sulfadoxine/pyrimethamine.

Of the 27 further SP samples collected in this study, one showed chippings upon visual inspection (hereafter called sample Z). Otherwise, all samples passed visual inspection, as well as TLC analysis and disintegration testing, performed according to the Minilab manual, and also testing of the uniformity of mass of dosage units.

Ten samples were subjected to a full pharmacopeial analysis according to the methods of the United States Pharmacopeia (USP) in the WHO-prequalified drug quality control laboratory of the Mission for Essential Drugs and Supplies (MEDS) in Nairobi, Kenya. These included the sample containing a mixture of tablets of types X and Y, the sample Z showing chippings, and eight further, randomly selected samples. While authentic sulfadoxine and pyrimethamine standards showed HPLC retention times of 8.24 and 3.26 min, respectively, tablets of type X showed a peak at 3.43 min (paracetamol), and tablets of type Y showed peaks at 9.07 min (sulfamethoxazole) and 2.11 min (trimethoprim), proving that these tablets did not contain the declared active ingredients. Using authentic standards for paracetamol and co-trimoxazole and the appropriated USP methods for these drugs, the tablets of type X and type Y were confirmed to represent paracetamol 500 mg tablets and co-trimoxazole 480 mg tablets.

The sample with chippings (sample Z) failed pharmacopeial analysis both for sulfadoxine content (71.8 % of declared content) and for dissolution of sulfadoxine and pyrimethamine (55.5 % and 52 % dissolution in 30 min). One further sample failed dissolution for pyrimethamine (37.6 % dissolution in 30 min). Of the total of 10 samples subjected to analysis according to the USP, therefore seven passed the analysis in all aspects.