It happens to all of us. We’re getting a new method developed and validated and then it comes time to run our negative urines. And everything comes up as positive! There are peaks for our analytes of interest in every urine that we run! How is that possible? In this blog post, I’m going to discuss some of the more troublesome analytes that I’ve encountered as far as finding an actual blank sample and what I’ve done to try and fix the issue.
In my experience, some of the most problematic analytes to develop methods for are nicotine, cotinine, tobacco-specific nitrosamines (TSNAs) and all of their related compounds and metabolites. These compounds are everywhere. Secondhand smoke, and even third-hand smoke can make someone test positive for nicotine. We also can’t be sure that the manufacturers of our test tubes, extraction columns, LC or GC vials, and everything in between have made sure that no one smokes near the products or that smokers don’t work with the products themselves.
Recently, I was trying to develop a method for analyzing TSNAs using ISOLUTE SLE+ 1 mL columns. This is a difficult method to develop to begin with. When I finally got my LC-MS/MS method developed, the extraction added another degree of difficulty. I started out extracting samples that were 10 ng/mL and 100 ng/mL. I wanted to eventually get down to the sub-10 pg/mL level. So once I had my extraction method developed, I decided to see how low I could see with my method. When I ran my samples, I found that peaks were present for everything: the 5 pg/mL samples had peaks of the same abundance as the 1 ng/mL samples and the straight negative urine samples (without internal sample added) had massive abundances as well. My instrument sample blanks were negative so I knew it wasn’t instrument carryover (Dan discussed this in a previous blog post). I extracted a couple of times so I didn’t think it was sample carryover during the extraction. I was at a loss until I remembered that I was analyzing these compounds in urine and that nicotine is a huge issue for secondhand smoke contamination. Maybe TSNAs are as well?
My next step was to run water as my sample matrix. With these samples, I still saw some small peaks for the negative samples in my retention time windows. However, these peaks were much less than when I used donor urine. I then tried Urisub, a synthetic urine. Urisub looked similar to water. There were small peaks present for the negative samples but not nearly as bad as the donor urine that I had been using. Based on all of this, it looks like I may not be able to reliably achieve the sub-10 pg/mL level that I had hoped to without doing a back extraction calculation.
What analytes do you have issues finding blank samples for? Are there other synthetic matrices that you’ve found that work?