When performing Dissolution, the Limit of Detection and Limit of Quantitation come up in 2 different ways. One question I get is "Do I need to determine the LOD/LOQ for my dissolution method?" The other question involves not being sensitive enough to read dissolution samples due to a poor chromophore or low dose - and you can't get a high enough LOD/LOQ. Let's talk briefly about both of these.
When validating a dissolution method, do you need to find the LOD/LOQ for your method? Usually, the answer to this is no. Dissolution is reported as a whole percentage - so there isn't much value in showing linearity below 0.5% of your label claim. If you meet that and still have LOQ - then there isn't much need to continue to test lower until you reach that LOD/LOQ. That being said, if you do anticipate there may be a lower dose of your product coming in the future - then it might not hurt to try to get to 1% of that level. Even getting to the 0.5% limit is aggressive in my view for dissolution linearity since what really matters is quantifying the amount of drug expected in the 1st time point. If your target for release at the first time point is 30% for example - you may only need to demonstrate that the lower limit of detection is a bit below that level such as at least 20%.
If you are in the situation where you cannot get appropriate LOD/LOQ at the first time point, then that does become a challenge to work through. In these situations, you have 2 options - increase the sensitivity of the analysis or decrease the dissolution volume. Decreasing the dissolution volume is possible - and there are several choices to get lower dissolution volumes, and along with that a higher drug concentration. Some options that are available are 100, 200, and 250mL vessels for traditional Apparatus 1 and 2. The 100mL vessel can be used with as little as 40mL, allowing for over 10x increase in concentration. Other low volume options exist as well for Apparatus 3, 4, and 7 too. One type of Apparatus 7 (Agilent 400-DS) can get as low as a 3mL dissolution volume!
ICH mentions 3 procedures for the determinations of LOD (limit of detection) & LOQ (limit of quantitation) -
1. Visual evaluation
2. S/N ratio
3. Std. Deviation of response & Slope
- LOD = (3.3 x SD) / Slope
- LOQ = (10 x SD) / Slope
— Slope can be achieved from the linearity curve.
— SD can be achieved from:
a. Background noises of certain no. of blank runs (SD of noises in several blank injections)
b. Calibration/linearity curve covering LOD level conc. (when determining LOD), or LOQ level conc. (in case of LOQ determination)
After determinations, LOD & LOQ have to be proved accordingly.
During method development, a good technique could be to run linearity at 9/10 conc. level covering approximate LOD. Data from these runs would be used to determine LOD & LOQ. Then linearity curve could be prepared by taking 5 or 7 level conc. (as per SOP or protocol) from these data starting from the BDL (below detection limit) level (Reporting Threshold). Or fresh preparation can be used by covering BDL.
Note 1: Method range should be from BDL to at least 120% of spec. LOQ is NMT BDL.
Note 2: For LOD/LOQ determination from curve, LOD/LOQ level concentrations have to be covered.
As per ICH, the definition of BDL (Below Disregard Limit) / Reporting Threshold:
“𝘈 𝘭𝘪𝘮𝘪𝘵 𝘢𝘣𝘰𝘷𝘦 (>) 𝘸𝘩𝘪𝘤𝘩 𝘢 𝘥𝘦𝘨𝘳𝘢𝘥𝘢𝘵𝘪𝘰𝘯 𝘱𝘳𝘰𝘥𝘶𝘤𝘵 𝘴𝘩𝘰𝘶𝘭𝘥 𝘣𝘦 𝘳𝘦𝘱𝘰𝘳𝘵𝘦𝘥.”
[Note: I saw many cases where people get confused. Suppose BDL for a method = 0.05%, Impurity A result = 0.05%. Should one report it & calculate in Total Impurities? NO. It's to be disregarded.]
Reference:
- Q2(R1) Guideline
- Ken Boda (Dissolution specialist at Agilent)
