TOP243RN field failure

We are having reliability issues with a design using TOP243RN as the main switcher on a flyback auxiliary converter. In all instances of the field failures, we've found the TOP243RN drain to source shorted and a fuse upstream opening. Replacing the fuse and the TOP243RN allows the converter to operate normally. Comparison of waveforms during the failure analysis showed no differences in operational modes between the failed unit and a control unit once TOP243RN and the fuse were replaced.

I've monitored the drain to source voltage and it does not exceed 540V throughout all operational conditions. Peak drain current is far below the maximum rating of 1.44A. Since it's part of the auxiliary converter, the input to the converter is pretty stable and is independent of unit operation (short circuit, input voltage fluctuations, etc).

Analyzing the unit date of return shows a component infant mortality-like curve with a majority of failures occurring within the first 1.5 years of operation, but we've also had a unit out in the field for over 3 years that have returned with the same failure mode. We're seeing a failure rate of about 150PPM.

I'm wondering if this can be attributed to infant mortality. Are there known reliability issues with these components? Anything I should investigate further?