Drug Modalities Explained: Small Molecules, Biologics, ADCs, Bispecifics, Cell and Gene Therapy, RNA, and Peptides

What Is a Drug Modality, and Why Does It Matter Before You Talk About the Disease?

A drug modality is the physical and chemical form a therapeutic takes, a small organic molecule, a protein, a strand of engineered RNA, a living cell, a viral vector, and it is a separate question from what disease is being treated or which biological target is being addressed. Two programs can aim at the identical target and diverge entirely in mechanism, cost, and commercial dynamics simply because one team chose a small molecule and the other chose an antibody.

This distinction matters early because modality constrains almost everything downstream: whether the drug can reach its target at all, how it is manufactured and at what cost, how it is dosed and how often, how durable its effect is, and how the resulting intellectual property and competitive landscape behave. A team that picks a modality by default, because it is what the founding lab happens to be equipped to make, is making a strategic decision without treating it as one. A team that picks deliberately, matching modality to the biology of the target and the realities of manufacturing and reimbursement, is doing something closer to company design.

Small-Molecule Drugs: How Do They Work, and Why Do They Still Dominate?

A small molecule is a low molecular weight organic compound, generally small enough to diffuse across cell membranes and bind a pocket on a protein target, whether that target sits on the cell surface, in the cytoplasm, or inside the nucleus. That ability to cross membranes, and in many cases cross the blood-brain barrier, is what makes small molecules the only practical modality for a large share of intracellular and central nervous system targets.

Manufacturing is chemical synthesis rather than biological production, which makes small molecules comparatively cheap, consistent, and scalable, and it is the main reason most small molecules can be formulated as oral pills rather than injections. Durability of effect is typically short, hours to a day, which usually means daily or multiple-times-daily dosing unless the molecule is chemically modified into a longer-acting or prodrug form.

The defining competitive dynamic of this modality is the patent cliff. Composition-of-matter protection runs for a period generally understood as roughly two decades from filing, and because the exact chemical structure can be synthesized by any competent chemist once that protection lapses, generic manufacturers can enter with bioequivalent copies that erode price sharply and quickly. In practice, years of that term are typically consumed by development and regulatory review before the drug ever reaches the market, so effective commercial exclusivity is usually shorter than the nominal term. This generic dynamic, well established in public policy discussion around drug pricing, is specific to small molecules and does not apply the same way to biologics.

Biologics and Monoclonal Antibodies: How Are They Different From Small Molecules?

Biologics are large, complex molecules, most commonly proteins, produced by living cells rather than synthesized chemically. Monoclonal antibodies, the largest category by revenue, are engineered to bind one specific target, usually a protein on the outside of a cell or circulating in blood, with a precision that is difficult to achieve with a small molecule. Because of their size, antibodies generally cannot cross into cells or across the blood-brain barrier, which confines them mostly to extracellular targets.

Manufacturing takes place in living cell cultures grown in bioreactors, commonly mammalian cell lines, followed by purification steps that are far more involved than chemical synthesis. This raises cost and introduces batch-to-batch variability that chemical manufacturing does not have to contend with, and it requires cold-chain handling that small molecules typically do not.

Delivery is almost always by injection or infusion, since the digestive tract would break the protein down before it could act, and durability is typically longer than a small molecule, days to weeks, aided by the way antibodies are recycled by the body rather than rapidly cleared.

The competitive dynamic differs meaningfully from small molecules. Because the manufacturing process itself shapes the final molecule, an idea often summarized as ’the process is the product,’ a competitor cannot simply copy the chemical structure once patents expire. Instead, follow-on competitors must develop biosimilars, which go through their own approval pathway demonstrating that the molecule is highly similar to, not identical to, the reference product, and price erosion after loss of exclusivity tends to be slower and shallower than the generic small-molecule pattern.

Antibody-Drug Conjugates: Combining a Guidance System With a Payload

An antibody-drug conjugate, or ADC, links a targeting antibody to a cytotoxic small-molecule payload through a chemical linker. The antibody portion binds an antigen on the surface of a target cell, the complex is taken into the cell, the linker is cleaved, and the payload is released where it can act, in principle sparing healthy tissue that lacks the target antigen.

This is one of the more manufacturing-complex modalities discussed here, because it requires three components, the antibody, the linker chemistry, and the payload, each developed and quality controlled separately, with the conjugation step itself needing to be highly reproducible from batch to batch. Delivery is by infusion, and dosing intervals tend to resemble those of biologics, generally weeks apart.

The intellectual property picture is correspondingly layered: the antibody, the linker chemistry, and the payload can each be separately patented, which gives a well-executed ADC a dense competitive moat but also frequently means the constituent technologies are owned by different parties, which is why licensing deals between smaller biotechs holding one piece of the puzzle and larger companies holding another are common in this space.

Bispecific Antibodies: One Molecule, Two Targets

A bispecific antibody is engineered to bind two different targets simultaneously, rather than the single target a conventional monoclonal antibody recognizes. A widely used design forces a physical connection between an immune T cell and a diseased cell, redirecting the immune system to attack independent of the T cell’s own native targeting, while other bispecific designs block two disease pathways at once with a single molecule.

Manufacturing is more difficult than for a standard antibody because correctly pairing two different sets of antibody chains, avoiding unwanted mismatched combinations, requires specialized protein engineering platforms rather than standard antibody production methods. Delivery is by infusion or subcutaneous injection, and because engaging the immune system this directly can provoke a strong inflammatory response, early dosing is sometimes staged more cautiously before settling into a longer maintenance interval.

Competitively, the underlying engineering platform used to force correct chain pairing is frequently patented and licensed as a standalone technology, so which bispecific platform a company has rights to can matter as much as the specific target pair it has chosen.

Cell Therapy, Including CAR-T: Turning Living Cells Into the Drug

Cell therapy uses living cells as the therapeutic agent itself. In CAR-T therapy, a patient’s own T cells, or in some newer approaches donor-derived T cells, are collected, engineered outside the body to express a chimeric antigen receptor that recognizes a disease target, expanded, and reinfused. Because the cells are alive, they can recognize, kill, and in some cases persist and continue expanding inside the patient long after infusion, which is a fundamentally different concept from a fixed, repeatedly administered dose.

Approved use to date has concentrated on blood cancers, with solid tumor and autoimmune applications still in earlier stages of clinical exploration. Manufacturing is among the most operationally complex of any modality covered here: an autologous approach requires a dedicated manufacturing slot for each individual patient, spanning collection, genetic engineering, expansion, and quality release, with turnaround measured in weeks and chain-of-custody logistics that are as central to the product as the biology itself.

Delivery is by infusion, typically after a course of chemotherapy intended to make room for the engineered cells to expand. The durability proposition can be substantial when it works, a single treatment producing a lasting remission in some patients rather than a chronic, repeated therapy, though outcomes vary and relapse remains a real risk. Competitively, patents cover the receptor construct, the engineering vector, and the manufacturing process, and a major current frontier is allogeneic, or donor-derived, cell therapy, which aims to remove the per-patient manufacturing bottleneck that defines the autologous approach.

Gene Therapy: Changing the Instructions Rather Than Managing the Downstream Effect

Gene therapy delivers genetic material into a patient’s cells, commonly a functional copy of a gene, a gene-editing system, or a construct designed to silence a gene, usually carried by a viral vector such as an adeno-associated virus or a lentivirus, though non-viral delivery approaches also exist. The intent is to address a genetic root cause directly rather than manage its downstream biological consequences, which is why the modality has concentrated first on monogenic and rare diseases where a single, well-characterized gene defect drives the condition.

Vector manufacturing is highly specialized, capacity constrained across the industry, and difficult to standardize, with potency assays and quality control that remain an active area of technical development rather than a solved problem. Delivery is frequently a single administration, whether by infusion, direct injection into the target organ or tissue, or into the spinal fluid, depending on where the affected tissue sits.

The durability promise is among the most notable in the field, a single treatment intended to produce effects lasting years, though because the modality is still relatively young, long-term durability data continues to accumulate for many programs rather than being fully established. Competitively, capsid and vector engineering, delivery technology, and the specific genetic construct are each separately patentable, but in practice, vector manufacturing capacity and supply agreements are frequently the real bottleneck to competition, arguably more so than patents alone.

RNA Therapeutics: How Do mRNA, siRNA, and Antisense Oligonucleotides Differ?

RNA therapeutics act at the level of genetic instructions rather than the finished protein or the DNA itself, and the three major subtypes work in distinct ways. Messenger RNA, or mRNA, delivers instructions that a patient’s own cells translate into a protein, typically packaged in a lipid nanoparticle to protect it and aid cellular uptake. Small interfering RNA, or siRNA, works through the cell’s natural RNA interference machinery to degrade a specific messenger RNA, silencing production of a chosen target protein. Antisense oligonucleotides, or ASOs, are short synthetic strands that bind a complementary RNA sequence directly to block, degrade, or alter the splicing of that transcript.

Use cases follow from these mechanisms: mRNA for vaccines and, increasingly, protein replacement; siRNA and ASOs for silencing disease-driving genes, with meaningful traction to date in liver-directed disease and a growing body of work in neuromuscular and central nervous system conditions. Manufacturing relies on chemical or enzymatic synthesis rather than cell culture, which tends to be more standardized and faster to scale than antibody or cell-based production, though formulating mRNA into a stable lipid nanoparticle adds its own layer of complexity.

Delivery is by injection across the class, subcutaneous for most siRNA and ASO programs, intramuscular for mRNA vaccines, and reaching tissues beyond the liver in a targeted way remains one of the central open technical challenges for the field as a whole. Durability can be a genuine differentiator: chemically stabilized siRNA conjugates have achieved dosing intervals stretching to several months in some approved uses, a real advantage over the daily or weekly dosing typical of many small molecules and biologics. Intellectual property here centers on the specific chemical modifications to the nucleotides and on delivery or conjugation platforms, which tend to be concentrated in a small number of specialist companies, making platform licensing a central part of the competitive picture.

Peptide Therapeutics: Positioned Between Small Molecules and Biologics

Peptides are short chains of amino acids, larger and generally more target-specific than small molecules, but structurally simpler and easier to manufacture than a full antibody. Mechanistically, many peptides mimic or block a natural hormone or signaling molecule, giving them a specificity closer to a biologic, though most still require injection because digestive enzymes break peptide bonds down before an oral dose could act, a limitation newer formulations using absorption enhancers are beginning to work around for select programs.

Manufacturing is chemical, generally via solid-phase peptide synthesis, rather than requiring a living cell line, which makes peptides more scalable and consistent than cell-culture biologics, even as cost can still be meaningful for larger or more complex peptides. Use cases concentrate in metabolic disease and endocrinology, with a growing presence in oncology.

Durability without chemical modification tends to be short, but techniques such as attaching a fatty acid chain can extend dosing intervals out to weekly administration in some cases. Competitively, peptides sit in a hybrid position: composition-of-matter patents apply as with small molecules, but because peptides can often be reverse engineered and reproduced once patents lapse, follow-on competition can resemble either the generic small-molecule pattern or something closer to the biosimilar pattern, depending on the complexity of the specific peptide and its manufacturing process.

How the Modalities Compare at a Glance

Laid side by side, a pattern emerges: modalities that are chemically simpler to manufacture, small molecules and peptides, tend to trade lower cost and easier scale for shorter durability and a more predictable, faster-eroding competitive cliff. Modalities that are biologically complex to manufacture, cell therapy and gene therapy in particular, trade higher cost and manufacturing risk for the possibility of a single treatment with a lasting or even one-time effect. Biologics, ADCs, bispecifics, and RNA therapeutics occupy the middle ground in different ways, each with its own delivery constraints and its own flavor of intellectual property dynamics.

• Small molecule: oral, cheap chemical synthesis, short dosing intervals, generic patent cliff
• Biologic and monoclonal antibody: injectable, cell-culture manufacturing, days-to-weeks durability, biosimilar competition
• Antibody-drug conjugate: infusion, three-part manufacturing complexity, weeks durability, layered multi-party IP
• Bispecific antibody: infusion or injection, specialized chain-pairing manufacturing, variable dosing, platform-licensed IP
• Cell therapy and CAR-T: infusion, per-patient manufacturing logistics, potentially durable single treatment, construct and process patents
• Gene therapy: single administration via vector, capacity-constrained manufacturing, intended long-term or one-time effect, vector and construct patents
• RNA therapeutics: injectable, chemical synthesis with formulation complexity, durability ranging from weeks to months, chemistry and delivery platform IP
• Peptide: mostly injectable with emerging oral options, chemical synthesis, durability extendable via modification, hybrid generic and biosimilar-style competition

Why Is Modality Choice a First-Order Decision for a University Spinout?

For a spinout emerging from academic research, the modality is often set implicitly by whatever the founding lab happened to be equipped to work with, a chemistry lab defaults toward small molecules, a protein engineering lab toward antibodies, a gene-editing lab toward gene therapy. That default is worth questioning explicitly rather than inheriting, because the modality decision, more than almost any other early choice, determines whether the science can actually become a product.

Translatability is the first question: does the biology of the target even permit the modality under consideration. An intracellular target with no accessible surface epitope generally rules out an antibody-based approach outright, regardless of how compelling the underlying biology is, while a target requiring long-lasting gene silencing may be poorly served by a small molecule that must be dosed daily.

Manufacturability is the second question, and it is where many academic programs are weakest, since a process that works at bench scale in a university lab often needs substantial redevelopment, sometimes a wholesale change in production method, to be manufactured reproducibly at clinical and commercial scale. Cost of goods follows directly from manufacturing complexity: a modality with a favorable mechanism can still be commercially unviable if its cost to produce leaves little room relative to what payers will reimburse, which is a more binding constraint for cell and gene therapy than for small molecules or peptides.

Finally, competitive crowding by modality matters as much as crowding by target. A target that looks open may already have several well-funded programs pursuing it with a particular modality, while a different modality against the same target may be comparatively uncontested, or may be blocked by a small number of parties holding the key platform patents. A pre-seed investor evaluating a university spinout is, in effect, asking whether the founding team chose its modality deliberately against these four questions, translatability, manufacturability, cost of goods, and competitive crowding, or backed into it because it was simply what the lab already knew how to make. That distinction is frequently visible well before there is any clinical data to evaluate, which is exactly why it belongs at the center of early diligence rather than at the end of it. This article is offered for general education about how drug modalities work and is not investment advice or an offer to invest in any fund or security.