What I would do to approach this problem:
1. Start from population, estimate the amount of population with cancer. I would segment by age because I believe that cancer is much more prevalent in mature ages
2. I probably would've missed the diagnosed rate actually.
3. I would calculate % of people treated by immunotherapy, segmenting by income level since people with lower income will less likely be treated
4. Estimate amount of therapy one person require, probably by frequency (once a week?) and $/week (idk, 1k$?)
5. Calculate the final results
Let me know your thoughts