Some considerations when picking the type of drug
In addition to what you are targeting, you also need to selection how you will target it - what drug class is the best for achieving your goal?
Your choice of drug class impacts a number of other factors and choices relevant to your drug from its clinical development to its eventual hopeful market approval and commercialization strategy.
The major drug types are summarized here. I’ve highlighted some variables that I usually go to first when thinking this problem through.
Is the disease caused by a loss- or gain-of-function in the target?
Loss-of-function (LoF) mutations are mutations where a gene or protein is inactivated or otherwise dysfunctional, and you want to turn it back on. Gain-of-function (GoF) is when a target is overactive or functioning in an undesirable way and you want to silence it.
LoF challenging to drug with small molecule as there is no protein to inhibit. Therefore, therapeutic strategies here attempt to replace or repair the lost gene, for example by increasing that specific protein or delivering a replacement. Gene therapy can work particularly well for this, depending on other variables; small molecules can sometimes be used by for example inhibiting and inhibitor of your LoF target, therefore increasing its prevalence. SSRIs are an example of this - they block the enzymes which degrade serotonin, allowing the existing serotonin to have more effect.
GoF is more traditionally druggable - small molecules and protein drugs can block the binding site of a protein, an antibody can target a protein or cell for degradation, RNAi can silence a gene.
Where is the target of the drug located?
The location of your target is very important when considering how you will drug it.
For a drug to reach an intracellular target, it must first cross the plasma membrane. The plasma membrane is a lipid bilayer and generally impermeable to large molecules (proteins, monoclonal antibodies). Small molecules are usually able to cross the plasma membrane; larger molecules usually need to enter via a channel or a transporter. RNAi, small molecules, and gene therapy are generally adept at hitting intracellular targets.
Protein drugs such as antibodies generally cannot cross the membrane. This can be beneficial if your target is extracellular (e.g., a specific receptor).
Does the disease affect all tissues or only certain organs?
A small molecule drug will be biologically active in most tissues. Therefore, you need to consider whether you want your drug to have biological activity in all these tissues and what the potential impact (positive and negative) of this could be.
Some drugs can be locally delivered (e.g., antibody to the eye, local small molecule injection into joint) - this can increase their efficacy if this is the only tissue that needs the medicine.
Gene therapies and RNAi generally struggle to go into tissues besides the liver and eye, although this is an area of rapid work.
(This is kind of a false questions as most diseases, even if technically located in one organ, are usually systematically influenced and driven)
Is the tissue behind the blood-brain or blood-retina barrier?
The brain and the eye have their own micro-environments, facilitated by specific cells that regulate what can enter them.
Generally, systemically delivered drugs do not cross the blood-brain/retina-barrier unless specifically optimized to. Or, they may cross, but there will inadvertently be a higher systemic pharmaceutical dose than CNS dose.
Cost tolerance & cost of manufacturing
Small molecules can usually be manufactured cheaply. Protein drugs will generally always be more expensive than small molecules.
Gene therapies, cell therapies, and RNAi are the most expensive drugs on the market. This is partially because of the expense of manufacturing, the economics (these drugs are currently designed for very small patient populations with no other therapeutic options), and the extra-long history of development. These types of drugs also have much more specialized manufacturing.
If you choose a more expensive drug class, but another company manages to develop a similarly-efficacious drug using a cheaper type of drug, you may not be able to win a price (or dosing convenience) war.
Drug safety
Different types of drugs have different safety profiles. In diseases where if unchecked can significantly harm or kill the patient, side effects can be more tolerated. In other diseases like high-blood pressure, side effects beyond the most minor are not tolerated.
Antibodies are designed to only bind specific proteins and do not enter cells, so they are generally less likely to have safety issues due to binding other proteins. Toxicity generally comes from the action of the antibody on the target.
Small molecules can bind less specifically and therefore have increased risk of off-target binding.
Cell & gene therapy are incipient but have shown increased immunogenicity in some uses (e.g., CAR-T carries a risk of cytokine release syndrome, an extreme immune reaction that can be deadly; gene therapy may not be dosable multiple times due to acquired immune reaction to the AAV capsid).
How patients take the drug
Oral delivery is generally not easily achieved for any drug class besides small molecule.
Pharmacokinetic considerations
Antibodies have a long half life and therefore only need to be dosed every 2-6 weeks. Small molecules are usually dosed on a daily basis. Gene & cell therapy are currently considered one-time therapeutics for most uses. Proteins often have a very short half-life unless modified.
Economics
A small molecule, biomolecule, or antibody drug can be dosed regularly, providing recurring revenue for your company. Cell & gene therapies are usually only dosed once or a few times, requiring the full financial value of the drug to be realized up front. This is a partial reason as to why these types of drugs are the most expensive. Additionally, the manufacturing and materials cost for everything but small molecules can be very expensive, eating into the potential margins.