If two injections of the same material result in two different results, then the analysis must stop. Results are invalid as clearly there is a problem with the method, instrument, sample and/or procedure used.
BTW: Two samples are not enough statistically to draw any conclusions from. Depending on the requirement, 3, 5 or 7 would be better.
Many reasons exist for one result to be "high" (area counts). A common reason for this is caused by a worn injector resulting in carryover. Maybe some instrument maintenance is in order. Flushing down an HPLC system is not the same thing as carrying out regular maintenance. The injector rotary valve seals and surfaces wear out over time. They trap sample material in the valve. This results in extra sample eluting out during an analysis (or wash step). You can read more about how to troubleshoot them in this free article, "
Carry-Over (Carryover) Contamination in HPLC and LC-MS Systems";
https://hplctips.blogspot.com/2015/02/c ... on-in.html