>>12333667The field of pharmacogenomics is actually concerned with this question, although it's a lot more nuanced than "hurr drug work for aryans but not niggers hurr". It mainly explores how an individual's genome affects how they respond to drugs, and how those drugs are metabolised in the body.
For example, take codeine, for pain relief (see
https://www.ncbi.nlm.nih.gov/books/NBK100662/). Codeine is actually a prodrug for morphine, and must be converted before it can even take effect. One of the steps for converting codeine involves the enzyme CYP2D6, in the liver.
A subset of individuals may carry more than two functional copies of the gene to make CYP2D6, meaning they overproduce CYP2D6. These individuals are known as "ultrarapid metabolisers", because their mutation allows them to convert codeine to morphine more rapidly and completely.
Thus, after taking codeine, at any single point in time, they will have much more morphine circulating in their system, as opposed to an individual with only two functioning copies of CYP2D6. This pretty much puts them at greater risk of overdose every time they take codeine.
There are many more examples of genetics affecting how treatments work. While the example I used was a small molecule, antibody drugs are also affected.