There's a few things that I notice:
1. The amplifier has no feedback so there is nothing to set the DC value of the output, hence it's heading off into the rail.
2. The input common mode voltage (0.9 V) is probably too high for the input stage.
3. The input signal of 1 MHz is a bit high for a first "proof of life" sim in my opinion
Personally, I would start with the amplifier in a buffer configuration (let's hope it's stable) : connect output to input negative. And input positive with a pulse at 1 kHz, 0.45 V amplitude, and a DC offset of 0.45 V