Spectral color

Assessing Color Changes in Stability Studies Using Spectrophotometric Chromaticity Measurements Versus Visual Inspection

In this study evaluating the color change of two different drug formulations under stress conditions, assessed by visual examination and spectrophotometric measurements, the results showed that spectrophotometric analysis was consistently able to detect color change long before examination. visual, especially when the changes were subtle (paracetamol solutions).

For the study, the two drug formulations that were chosen (paracetamol solutions and parenteral nutrition solutions) are well known for their propensity to change color under stress. Paracetamol is a widely used drug but it is sensitive to photodegradation. This phenomenon appears to be caused by hydrolysis and oxidation accelerated by exposure to radiation. In particular, the decomposition of paracetamol into quinonimine and related compound leads to a gradual color change from pink to brown by oxidative degradation in solid paracetamol, while its degradation when solubilized appears to be caused by a catalytic hydrolysis reaction acidic and basic and leads to what has been described as initially a pink color which darkens to brown over time16,17,18.

Amino acids containing SNPs are also known to be susceptible to changing color and becoming more brown over time, especially at room temperature. This phenomenon is generally attributed to the Maillard reaction and the degradation of the amino acids they contain, in particular lysine, glycine and methionine.19. Nevertheless, Yailian et al. showed that the SNP initially became more yellow than brown, and linked this to the increased presence of cysteine, which is the oxidized dimer form I-cysteine20. This information validates the use of the European Pharmacopoeia color scheme B (brown) and Y (yellow) as references for visual examination for paracetamol and PNS formulations respectively.

As expected, both drug formulations changed color during the study when exposed to their stress condition (light for the paracetamol solutions and room temperature for the SNP, which remained sterile for the stress conditions). storage). In both cases (stressed drug formulations), spectrophotometric analysis objectified a change in color and luminosity before it was detected by visual observation, with a difference of 2 weeks for paracetamol solutions (color change detected on D14 versus D28 at the best) and 1 week for the PNS (change in color detected on day 7 versus day 14), for spectrophotometer analysis and visual examination respectively. Interestingly, when the solutions were stored under optimal storage conditions (in the dark for the paracetamol solutions and at 5°C for the PNS), the difference in detection capacity was even more obvious. When spectrophotometric analysis detected a change after 14 and 7 days (for paracetamol solutions and PNS respectively), visual examination either did not notice a change (paracetamol solutions) or did not notice the change only after 46 days of storage. These results are consistent with the calculated ∆E value, which expresses the difference in color perception. It has been proposed that when ∆E 21although this general simplification may not be exact for all colors, variations having been observed depending on the original chromaticity during studies of forced degradation22. Also, other authors have proposed other limits: Faghihi et al.23 indicates that ΔE 3.3: easily observable with the naked eye, while Gupta et al.24 uses even wider acceptability thresholds (ΔE > 3.7 – easily visible difference, ΔE between 3.7 and 1 acceptable difference, ΔE 0.5 for a* and b*). When exposed to light, the ΔE values ​​were 1.381 and 2.653 on days 28 and 39, respectively, and it was on these days that visual examination just started to notice a difference. This result was also valid for SNP: ΔE values ​​were 1.39 at 39 days and 2.02 after 46 days at 5°C, and the color change was detected by observers at day 46 (but not on day 39). However, for the 25°C storage condition, ΔE was 2.8 after 7 days, thus suggesting an easily detectable color change, but the change was not detected at day 7 (p=0.68 and 0.14 for a* and b*, respectively) but only on day 14 (p

The increased variability observed for visual examination results could be explained by several factors, one of which could be differences in perception. This variability is inherent in any visual examination. The use of a double-blind, randomized system meant that the observers were unable to identify any of the tubes, thus ensuring that they were not influenced in any way. However, the correspondence of the visual observation with comparable values ​​of a*, b* and L* was limited by the European Pharmacopoeia reference range solutions B and Y, as there were only 9 and 7 points respectively in color schemes. Observers who saw colors intermediate between two color levels (eg between Y6 and Y5) had to choose a level from the range of solutions. This evaluation method could have limited the personal perception of each observer and constitutes one of the limitations of the study. Also, although the visual examination was standardized (visual examinations carried out under the illumination of an LED lamp diffusing a white light of 4000 K), the ambient lighting of the room in which the study was carried out was not standardized. Differences in ambient light (related to different climatic conditions, for example) could also impact technicians’ perception of colors and reduce their ability to accurately assign the color of solutions. Unless this parameter is carefully controlled, it is another limitation of any visual examination performed for color assessment during stability studies. A sample size of three (as was used in this work) is in common use in several scientific fields, including practical stability studies (see Bardin et al.25 and the recommendations issued jointly by the French Society of Clinical Pharmacy (SFPC) and the European Society for Hospital Pharmaceutical Technologies (GERPAC)26). However, like any sampling, it can lead to bias. It is possible that the number of samples was insufficient to allow an ideal assessment of the color by the examiners (perhaps also explaining the variability observed for the results of the visual examination), but as this is representative of real situations , this only serves to reinforce the use case of spectrophotometric colorimetric analysis in practical stability studies. In this study, CIELab formulas were used, as opposed to CIEDE2000 color difference formulas. Using the latter could have resulted in a better fit than with the CIELAB formulas, but it seems unlikely that the differences would have an impact on the interpretation of our results. Indeed, by comparing the CIELab and CIEDE2000 color difference formulas, the work published by Gomez-Polo et al.27 indicates that although the linear correlation coefficients estimating color differences with the CIELAB and CIEDE2000 formulas compared to participants’ perceived colors were higher with CIEDE2000 than with CIELAB, they remained equally low (0.289 and 0.176, respectively). In addition, the CEEDE2000 is considerably more complex to use28and could lead to misinterpretations if used in practical drug stability testing.

Drug stability studies typically include a visual examination criterion, during which the observer should note any changes in appearance, and look for haziness, the presence of visible particles, and a change in color.11, 25, 26, but do not include any definition of a maximum permitted color change possible due to the limitations of usual visual examination. However, few authors include spectrophotometric analysis of potential color change and rely on subjective visual assessments, thus potentially missing early signs of instabilities. Indeed, to our knowledge, it seems that only three recent works on the stability of drugs have included this analysis in their study. Win et al. proved that the degradation of pemetrexed was related to a color change towards yellowish solutions, confirmed by increasing b* values, and that the addition of different antioxidants to the formulation not only decreased various degradation products but also limited the variation of the values ​​of b*29. Chennell et al. also showed that the evolution of the color towards red measured by colorimetry of the solutions of amphotericin B occurred in parallel with a decrease in the concentrations of amphotericin B. However, the appearance of a degradation product (evaluated by a stability indicator method) does not induce any color change for atropine solutions stored at 25°C30. The colorimetric analysis could also indicate the presence of other compounds, for example as toxins31. Overall, spectrophotometric colorimetric assessment of color change is another tool, simple to implement, that could help in assessing the stability of drugs during stability testing.