If I understand the data correctly, the lead level decreased with time. Again, it appears to have decreased to a level where lead remediation was no longer required.
Please pardon the suspicion, but this is very "convenient". Lead remediation can be expensive. If the original sample is available, retesting the original sample by another lab would probably be wise. Look out if the original sample was "lost" (again, this would be very convenient).
As for how the level can change, a different sample from a slightly different location could easily have lower levels. If one sample was nearer to a lead based solder joint, for example. How was the second sample pulled and what were the differences?
There are many other variations (external contamination, dirty glassware in testing, etc.). Starting with the sample location is probably best.