Hello @Harsh Thakor ,
I discussed this scenario with the Product group team and their response is as below:
Because NAT gateway IPs cannot be shared across regions, the best solution for the proposed set up is to disassociate the public IP prefix from each NAT gateway in each respective region and to assign a single public IP address to each instead. This limit of 50K connections on NAT gateway refers to the number of connections per IP going to the same destination endpoint that are supported at any given time on the NAT gateway. The key here is not how many connections can be made in a day, but rather how many connections the customer has open on their NAT gateway at any given time. So long as the customer does not have more than 50k connections open at the same time going to the same destination endpoint, a single IP address on their NAT gateway should be able to support the number of requests made.
If you do not expect more than 50K connections open on your NAT gateway at any given time in a day, then you can remove the Public IP prefixes and use Public IPs instead which would reduce the total to 2 IPs.
If you expect more than 50K connections open on your NAT gateway at any given time in a day, then as suggested by @msrini-MSFT below, the best option is to go with NAT gateway and reduce the IP of the DR setup by 1 with a total of 3 IPs in your setup.
----------------------------------------------------------------------------------------------------------------
Please "Accept the answer" below if the information helped you.