Division of Pharmaceutical Analysis, Center for Drug Evaluation and Research, U.S. Food and Drug Administration , 645 South Newstead Avenue, Saint Louis, Missouri 63110, United States.
Anal Chem. 2016 May 3;88(9):4706-13. doi: 10.1021/acs.analchem.5b04636. Epub 2016 Apr 13.
Potential infiltration of counterfeit drug products-containing the wrong or no active pharmaceutical ingredient (API)-into the bona fide drug supply poses a significant threat to consumers worldwide. Raman spectroscopy offers a rapid, nondestructive avenue to screen a high throughput of samples. Traditional qualitative Raman identification is typically done with spectral correlation methods that compare the spectrum of a reference sample to an unknown. This is often effective for pure materials but is quite challenging when dealing with drug products that contain different formulations of active and inactive ingredients. Typically, reliable identification of drug products using common spectral correlation algorithms can only be made if the specific product under study is present in the library of reference spectra, thereby limiting the scope of products that can be screened. In this paper, we introduce the concept of the Raman barcode for identification of drug products by comparing the known peaks in the API reference spectrum to the peaks present in the finished drug product under study. This method requires the transformation of the Raman spectra of both API and finished drug products into a barcode representation by assigning zero intensity to every spectral frequency except the frequencies that correspond to Raman peaks. By comparing the percentage of nonzero overlap between the expected API barcode and finished drug product barcode, the identity of API present can be confirmed. In this study, 18 approved finished drug products and nine simulated counterfeits were successfully identified with 100% accuracy utilizing this method.
假冒药品产品（含有错误或没有活性药物成分的药品）渗透到合法药品供应中，对全球消费者构成了重大威胁。拉曼光谱提供了一种快速、非破坏性的方法来筛选高通量的样品。传统的定性拉曼识别通常采用光谱相关方法，即将参考样品的光谱与未知样品的光谱进行比较。这对于纯物质通常是有效的，但在处理含有不同配方的活性和非活性成分的药物产品时，就极具挑战性。通常情况下，只有在研究中特定的产品存在于参考光谱库中时，才能使用常见的光谱相关算法对药物产品进行可靠的识别，从而限制了可以筛选的产品范围。在本文中，我们通过将 API 参考光谱中的已知峰与研究中的成品药物中的峰进行比较，引入了药物产品识别的拉曼条码概念。该方法需要通过将 API 和成品药物的拉曼光谱转换为条码表示，将除对应于拉曼峰的光谱频率之外的所有光谱频率的强度赋值为零。通过比较预期 API 条码和成品药物条码之间非零重叠的百分比，可以确认 API 的存在。在这项研究中，利用该方法成功地识别了 18 种已批准的成品药物和 9 种模拟的假冒药物，准确率达到 100%。
