Which startup will solve cancer?
In our biotechnology market deck, you will find everything you need to understand the market
SUMMARY
A2 Biotherapeutics currently has the strongest private therapeutic claim, but no startup is close to solving cancer at population scale.
A real solution would need to cause a large and lasting fall in deaths across several common cancers. One dramatic remission or a drug for a rare mutation can prove that a mechanism works, but it cannot meet that standard.
The awkward truth is that prevention may save more lives than any startup treatment now in development. Tobacco control, vaccination, healthier environments and existing screening could prevent a large share of cancers before a new drug is ever needed.
Early detection could also outrank treatment in eventual impact, although the evidence is not settled. GRAIL’s large NHS trial found more early cancers and fewer stage IV diagnoses, but it missed its main endpoint and has not yet shown a mortality benefit.
The strongest benchmark for a therapeutic platform comes from Revolution Medicines. Daraxonrasib roughly doubled median survival in previously treated metastatic pancreatic cancer, showing the level of controlled human evidence that private startups still lack.
Solid tumors remain the central test. Cell therapies have changed several blood cancers, yet common lung, colorectal, pancreatic and breast tumors still block immune cells, vary from cell to cell and often share targets with healthy tissue.
A2 Bio’s Tmod system is unusually interesting because it checks both for a tumor target and for the absence of a healthy-cell marker. That logic could open targets that would otherwise be too dangerous to attack.
The company’s lead rests on one remarkable patient: a confirmed complete response in metastatic lung cancer, followed later by an isolated central-nervous-system relapse. The result is real and clinically unusual. It is also still one patient.
A2 Bio’s biggest obstacle may be manufacturing rather than biology. Its current product is made separately for each patient, takes weeks and requires specialist treatment, which is difficult to reconcile with the scale of common cancers.
ArsenalBio could overtake A2 Bio with a broader programmable-cell platform, Affini-T could gain ground by targeting KRAS, and Umoja could change the economics of the field through in-vivo CAR-T. None has yet published a stronger solid-tumor patient result.
The deciding evidence will be repetition: several durable responses at the same dose, activity across more than one tumor type, better control of brain metastases and a donor-derived or in-vivo product that can reach ordinary hospitals.
Our conclusion is blunt. A2 Biotherapeutics is the cancer startup to watch today, but solving cancer will require several technologies working together, and its current lead could disappear after the next meaningful clinical readout.
What would it mean for a startup to solve cancer?
A startup would count as solving cancer only if its technology caused a large and lasting fall in deaths across several common cancers.
One spectacular remission would fall well short of that standard. So would a treatment that works for a few thousand people carrying an unusual mutation. Both could become important medical advances, yet neither would transform cancer at population scale.
The latest global figures make the required scale clear. The American Cancer Society and the International Agency for Research on Cancer estimate that almost 21 million people developed cancer worldwide in 2024 and 9.8 million died from it. Annual diagnoses are projected to reach roughly 34.4 million by 2050 as populations grow and age.
A genuine solution would therefore need to help millions of people. It would probably combine several kinds of progress: fewer preventable cancers, earlier diagnoses, treatments that stop common cancer-driving mutations and immune therapies that clear advanced disease.
That definition favors platforms over isolated drugs. We are looking for technology that can be adapted across tumors, survive biological resistance and eventually reach far more patients than a bespoke experimental treatment.
Why can’t one startup cure every cancer?
No single startup can cure every cancer because two patients with the same diagnosis may have biologically different diseases.
Cancer begins through changes in DNA, but the relevant changes vary by patient. The National Cancer Institute also notes that two cells inside the same tumor can carry different genetic alterations. Additional changes appear as the cancer grows or encounters treatment.
A drug may kill the cells that depend on one mutation while leaving a smaller resistant group alive. Those survivors multiply, and the returning cancer may respond differently from the original tumor.
The problem becomes even harder after cancer spreads. A patient can have several metastatic sites with different mutations, immune environments and drug sensitivities. One treatment may control the liver tumors while failing in the brain or lungs.
Cancer will probably be beaten through a collection of tools rather than one universal medicine. The leading startup still needs some common leverage point, such as a recurring mutation, a shared immune mechanism or a platform that can be quickly redesigned for each patient.
This market map, featured in our biotechnology market deck, highlights top companies and startups in the biotechnology market
Is cancer actually becoming easier to beat?
Cancer is easier to survive today, although worldwide cases are growing too quickly for us to declare victory.
The latest American Cancer Society report found that the US cancer death rate had fallen 34% from its 1991 peak through 2023. That decline prevented an estimated 4.8 million deaths. Five-year survival across all cancers has also reached 70% in the United States.
Several changes produced that progress together. Smoking fell, screening caught more breast and colorectal cancers early, targeted treatments improved outcomes in cancers carrying specific mutations, and immunotherapy gave some patients durable control of previously lethal disease.
The global picture remains harsher. Nearly 9.8 million people died from cancer in 2024, and annual diagnoses are expected to rise by around two-thirds by 2050. A technology can work very well while the total number of patients continues climbing.
Cancer research is winning many individual battles. The overall burden is still expanding.
| Measure | Latest evidence | What it tells us |
|---|---|---|
| US cancer death rate | Down 34% since 1991 | Existing prevention, screening and treatments already work at scale |
| US deaths avoided | About 4.8 million | Progress has come from several advances working together |
| Global cancer deaths | About 9.8 million in 2024 | The remaining burden is enormous |
| Projected annual cases | About 34.4 million by 2050 | New technologies must outrun population growth and ageing |
Could prevention save more lives than any cancer startup?
Cancer prevention could save more lives than any treatment startup currently in development.
A recent World Health Organization and IARC analysis concluded that up to four in ten cancer cases worldwide could be prevented. The causes included tobacco, alcohol, excess body weight, ultraviolet radiation, air pollution and cancer-causing infections.
Vaccines already show what prevention can achieve. HPV vaccination can prevent most cervical cancers and several cancers of the throat and genitals. Hepatitis B vaccination lowers liver-cancer risk. Tobacco control has played a major role in the long fall in lung-cancer mortality.
Colorectal screening goes one step further by finding and removing precancerous polyps. The patient avoids both advanced cancer and the need for a future cure.
Most of this impact will come from public-health systems, vaccines, regulation and ordinary medical care. A startup could improve access or develop a new preventive vaccine, although it would struggle to own the entire solution commercially.
Today, the biggest unused opportunity in cancer remains applying prevention measures that already exist. The most exciting biotech platform may never match their potential reach.
If you want more recent data on this point, please see our latest biotechnology market report.
As this chart shows, and as featured in our biotechnology market deck, search interest in biotech has been trending upward
Could early cancer detection matter more than a cure?
Earlier cancer detection could prevent more deaths than many new treatments, provided that finding tumors sooner actually helps people live longer.
The idea sounds obvious. A small localized tumor can often be removed or destroyed, while cancer that has spread around the body becomes much harder to eliminate. The difficulty lies in proving that a blood test finds dangerous cancers early enough to change that outcome.
GRAIL’s NHS-Galleri trial has now provided the strongest test of this idea. More than 142,000 adults were randomly assigned to standard screening alone or standard screening plus three annual Galleri blood tests. The trial missed its primary target: combined stage III and IV diagnoses did not fall significantly.
The secondary results were more encouraging. Stage IV diagnoses fell 14% across the full study and 26% during the third screening round. Galleri also increased early-stage diagnoses by 16%, found four times as many cancers through screening and reduced emergency cancer presentations by 25%.
Those findings are promising, although deaths have not yet been shown to fall. Galleri also detected only 54.7% of the 12 cancers studied within a year of each blood draw, and sensitivity across all cancers was 30.7%.
So the latest large trial has moved multi-cancer blood testing forward without settling the argument. Earlier diagnosis is appearing. A survival benefit still has to follow.
Which cancer-detection startup is strongest today?
ClearNote Health currently has the strongest private-startup evidence for detecting one lethal cancer, while Harbinger Health is building the broader commercial platform.
ClearNote’s Avantect test focuses on people with an elevated risk of pancreatic cancer. Its updated validation involved 1,445 people and reported 82.6% sensitivity overall, 76.8% sensitivity for stage I and II disease, and 97.5% specificity. The test is commercially available, though it has not received FDA clearance or approval.
That focus is valuable. Pancreatic cancer is usually discovered too late for surgery, so a reliable test for high-risk patients could move more people into a potentially curable stage.
Harbinger Health is pursuing a different route. Its RESOLVE test is aimed at people who already have a suspicious scan, symptom or clinical finding. In an ASCO 2026 study of 209 such patients, 155 had cancer. Harbinger detected 85 of them and produced one false positive among the 54 people without cancer.
Harbinger is trying to shorten the uncertain period between suspicion and diagnosis rather than screening everyone. That may become a faster path into routine care, and the company recently raised $100 million to bring its first RESOLVE products toward commercial launch.
ClearNote wins today on disease-specific early-stage performance. Harbinger has the more flexible commercial strategy. GRAIL remains the evidence benchmark for the entire sector, although it is now a listed company rather than a private startup.
| Company | Main use | Strongest current evidence | Biggest missing proof |
|---|---|---|---|
| ClearNote Health | Detect pancreatic cancer in high-risk people | 76.8% sensitivity for stage I and II disease | Prospective evidence that testing increases curative surgery or survival |
| Harbinger Health | Help diagnose patients after cancer is suspected | Only one false positive among 54 non-cancer cases in a 209-person study | Larger real-world studies showing faster and better diagnosis |
| GRAIL | Screen broadly for more than 50 cancers | Randomized study of 142,250 adults | A demonstrated fall in cancer mortality |
If you want more recent data on this point, please see our latest biotechnology market report.
This chart, featured in our biotechnology market deck, illustrates yearly venture capital funding for biotechnology startups
Will an AI drug-discovery startup solve cancer?
AI drug-discovery startups will speed up cancer research, but none has shown that its models can repeatedly produce successful cancer medicines.
Xaira Therapeutics, Isomorphic Labs and Recursion can search biological data, identify possible targets and design molecules faster than traditional teams. Those abilities may shorten the early stages of drug development.
The hardest failures happen later. A molecule that looks excellent in a computer model can fail because it reaches the wrong tissue, harms healthy cells, works only in mice or loses effectiveness as tumors evolve.
The historical numbers are unforgiving. BIO studied 4,179 oncology programs that had entered Phase I and found that only 5.3% eventually reached approval. Solid-tumor programs performed even worse, with a 4.6% estimated chance of approval from Phase I.
AI could improve those odds, although we currently lack enough completed clinical programs to measure the difference. Recursion has computationally designed oncology drugs in human trials, while Xaira and Isomorphic Labs are still building much of their clinical evidence base.
These companies may become extremely valuable. Today, their cancer claim rests largely on a better research process rather than proven patient outcomes.
What does daraxonrasib tell us about the real cancer race?
Daraxonrasib shows that attacking a common cancer-driving mutation can produce a much stronger result than most startup platform stories.
Revolution Medicines designed daraxonrasib to block active RAS proteins. RAS mutations drive more than 90% of pancreatic cancers and also appear frequently in lung and colorectal tumors.
In the randomized Phase III RASolute 302 trial, patients with previously treated metastatic pancreatic cancer lived a median of 13.2 months on daraxonrasib, compared with 6.7 months on chemotherapy. The drug cut the risk of death by 60%, and the company is preparing regulatory submissions.
Doubling median survival in metastatic pancreatic cancer represents a serious clinical advance. Most patients still face eventual progression, so daraxonrasib is better described as a powerful treatment than a cure.
Revolution Medicines has been publicly traded for years, which removes it from our private-startup ranking. It provides a useful benchmark: a cancer platform earns its reputation when hundreds of patients live longer in a controlled trial.
Any private startup claiming broader potential currently has far less proof.
This chart, featured in our biotechnology market deck, looks at Vertex’s strategy in biotechnology
Why are solid tumors the biggest test for cancer startups?
Solid tumors are the biggest test for cancer startups because they cause most cancer deaths and have resisted the cell therapies that transformed several blood cancers.
CAR-T cells have produced long remissions and apparent cures in some leukemias, lymphomas and myelomas. These cancers offer relatively clear targets, and the engineered cells can reach malignant cells circulating in blood or living in bone marrow.
A solid tumor creates more obstacles. Immune cells must enter a dense mass, survive a hostile chemical environment and distinguish cancer from nearby healthy organs. Different cells within the tumor may also display different targets.
We now know that cell therapy can work in selected solid cancers. The FDA approved lifileucel, sold as Amtagvi, for advanced melanoma and Tecelra for a genetically defined group of patients with synovial sarcoma. These approvals proved that engineered or expanded immune cells can control solid tumors, though both treatments serve narrow populations and require demanding manufacturing.
No CAR-T therapy has yet gained approval for a common solid tumor such as lung, colorectal, pancreatic or breast cancer. That gap explains why the A2 Bio and ArsenalBio programs deserve attention.
A startup that repeatedly clears common solid tumors without damaging healthy tissue would open one of the largest remaining areas in oncology.
What makes A2 Biotherapeutics different?
A2 Biotherapeutics stands out because its engineered T cells check two biological clues before attacking a cell.
Most CAR-T therapies use one target. When the engineered cell sees that target, it attacks. This works well when cancer cells carry the target and essential healthy cells do not.
Solid tumors rarely provide such clean targets. Mesothelin, EGFR and other attractive proteins can also appear on normal tissue, creating a risk that a powerful treatment damages healthy organs.
A2 Bio’s Tmod system adds a blocker. Its lead therapy, A2B694, attacks cells carrying mesothelin but stops when it also sees HLA-A*02, a marker retained by the patient’s healthy cells. Eligible tumor cells have lost HLA-A*02, so the protective brake disappears inside the cancer.
In simple terms, the cell asks two questions: “Do I see the tumor target?” and “Do I see the marker that says this cell is healthy?” It attacks only when the first answer is yes and the second is no.
That design gives A2 Bio a practical way to use cancer targets that would otherwise be too dangerous. The company can also swap the activating target or add extra features while keeping the same basic safety logic.
This chart, featured in our biotechnology market deck, illustrates yearly funding for biotechnology startups
Has A2 Biotherapeutics actually worked in patients?
A2 Biotherapeutics has worked spectacularly in one lung-cancer patient, while the remaining clinical evidence is still very thin.
The EVEREST-2 study had enrolled 13 patients across six solid-tumor groups when A2 Bio presented its latest results. The group included colorectal, pancreatic, ovarian, lung and gastro-esophageal cancers, plus mesothelioma. Patients received several dose levels of A2B694.
One patient had metastatic non-small-cell lung cancer carrying KRAS G12V and STK11 mutations. The cancer had already progressed after carboplatin, pemetrexed and pembrolizumab.
A scan 90 days after treatment showed a complete response, and an independent review confirmed it at day 180. A PET scan soon afterward found no visible disease. The result was especially unusual because complete responses to CAR-T therapy are extremely rare in solid tumors and had not previously been reported in lung cancer.
Cancer later returned at one isolated site in the central nervous system. Disease outside the brain remained absent, and the engineered cells could still be detected in the patient’s blood after 15 months.
The early safety record was also encouraging. There were no dose-limiting toxicities, no treatment-related deaths and no patients leaving the study because of adverse events. One patient experienced grade 3 neurotoxicity, and another had mild cytokine-release syndrome while also receiving low-dose interleukin-2.
This patient gives A2 Bio a credible clinical foothold. The study has yet to establish a reliable response rate.
Is one complete response enough to make A2 Bio the leader?
One complete response gives A2 Bio the lead among private therapeutic startups, though the lead remains fragile.
The response was confirmed through several methods and occurred in a patient whose cancer had already resisted standard treatment. That makes it far more informative than a laboratory experiment or a mouse study.
The trial design still leaves enormous uncertainty. EVEREST-2 is an open-label Phase I study testing several doses in patients with different cancers. Only one patient in the reported group had lung cancer, so we cannot tell whether the response reflects the platform, the dose, the patient’s biology or a rare piece of good luck.
Early oncology results disappear regularly. BIO’s historical analysis found that only around one in twenty oncology programs entering Phase I eventually reached approval. A complete response improves our view of the mechanism, while leaving the product’s probability of success low.
Still, one patient is one patient. A2 Bio now needs repetition. Two or three more durable responses at a consistent dose would change the company’s position more than another financing round, partnership or preclinical presentation.
For now, A2 Bio leads on the quality of its best patient result rather than the quantity of evidence.
If you want more recent data on this point, please see our latest biotechnology market report.
This chart, featured in our biotechnology market deck, compares the main business model options for biotech platform companies
How many cancers could A2 Bio realistically treat?
A2 Bio could eventually treat several common solid tumors, although each therapy will reach only a selected subgroup of patients.
A2B694 requires three conditions. The patient must carry the correct inherited HLA configuration, the tumor must have lost HLA-A*02, and its cells must express enough mesothelin. Many patients will fail at least one of those tests.
The current EVEREST-2 study covers colorectal, pancreatic, ovarian, lung and gastro-esophageal cancers, along with mesothelioma. This gives the program breadth across organs, though biological eligibility will narrow the addressable population inside each cancer.
A2 Bio is already trying to expand beyond the original product. A2B543 uses the same mesothelin activator and HLA blocker while adding an IL-12 booster designed to strengthen the immune response around the tumor. The first patient has been treated, and dose escalation is underway.
Another program, A2B395, replaces mesothelin with EGFR and uses donor cells rather than cells collected from each patient. Its trial includes colorectal, lung, head-and-neck, triple-negative breast and kidney cancers.
The company’s reach will depend on whether this basic design keeps working after A2 Bio changes the tumor target. One successful mesothelin therapy would create a useful medicine. Several successful activators using the same blocker would validate a true platform.
Can A2 Bio make cell therapy cheap enough?
A2 Bio cannot become a large-scale cancer solution while its treatment remains slow, personalized and difficult to manufacture.
A2B694 is an autologous therapy. Doctors collect T cells from one patient, ship them to a manufacturing facility, modify and grow them, test the final product, and send the cells back for infusion.
The National Cancer Institute says the full CAR-T process currently takes around three to five weeks. Patients also need chemotherapy before infusion and close monitoring afterward because serious immune and neurological complications can occur.
That model has worked for specialized blood-cancer treatment, where current products can cost hundreds of thousands of dollars before hospital and follow-up expenses. It becomes much harder to imagine at the scale of lung, breast or colorectal cancer.
A2B395 is the company’s attempt to remove this bottleneck. It uses cells from healthy donors, creating the possibility of producing batches in advance and delivering them when needed.
Donor-derived cells bring their own problems. A patient’s immune system may reject them quickly, and the cells must be engineered so they do not attack the patient. A2 Bio has started clinical testing but has not disclosed evidence that A2B395 can persist or shrink tumors in people.
The personalized product may establish that the biology works. The donor-derived version will decide whether A2 Bio can ever reach a mass market.
This chart, featured in our biotechnology market deck, breaks down revenue across customer segments in the biotechnology market
Could ArsenalBio overtake A2 Bio?
ArsenalBio could overtake A2 Bio because its programmable T cells have a broader design, but the company currently lacks equally strong public patient data.
ArsenalBio builds engineered cells containing several biological instructions. AB-1015, developed for ovarian cancer, uses an AND gate that requires two tumor-related targets before the cell fully activates. AB-2100 uses another two-target system for clear-cell kidney cancer. A prostate-cancer program is also moving through the pipeline.
The company is trying to handle several solid-tumor problems inside one cell: target selection, safe activation, persistence and resistance to the suppressive environment around a tumor.
That architecture may eventually prove more flexible than A2 Bio’s activator-and-blocker system. ArsenalBio has also raised enough capital to run several clinical programs at once.
The missing piece is visible human efficacy. ArsenalBio has active Phase I studies, yet it has not publicly reported a patient result matching the lung-cancer response produced by A2 Bio.
ArsenalBio may have the more ambitious engineering platform. A2 Bio currently has the clinical lead. One meaningful response dataset could reverse the order.
If you want more recent data on this point, please see our latest biotechnology market report.
Could a KRAS startup beat A2 Bio?
A KRAS-focused startup could beat A2 Bio by reaching more patients with a target that directly drives several common cancers.
KRAS mutations appear frequently in pancreatic, colorectal and lung cancer. Tumors often depend on the altered protein for growth, making KRAS more fundamental than a surface marker that cancer cells may discard.
Affini-T Therapeutics is developing AFNT-211, a personalized T-cell therapy that recognizes the KRAS G12V mutation. The program includes advanced pancreatic, colorectal, lung and other solid tumors. It entered Phase I testing after the first patient was treated in 2024.
Eligibility is narrow. Patients need the KRAS G12V mutation and a matching HLA type. The cells must also reach solid tumors, stay active and avoid immune escape.
The wider KRAS field has gained credibility from daraxonrasib, the public-company program discussed earlier. Its Phase III survival benefit proves that broad RAS inhibition can change outcomes in pancreatic cancer.
A private KRAS startup could eventually offer greater reach than A2 Bio. Today, no private contender has released comparable human evidence.
This chart, featured in our biotechnology market deck, shows how at-home genetic testing technology has evolved over time
Could in-vivo CAR-T change the winner?
In-vivo CAR-T could change the entire ranking by turning cell therapy from a custom procedure into a standard infusion.
Conventional CAR-T treatment removes cells from the patient and engineers them in a factory. In-vivo technology sends genetic instructions directly into the body, where the patient’s own T cells are reprogrammed.
Umoja Biopharma is one of the strongest private companies pursuing this approach. The FDA has cleared its UB-VV400 program, an in-vivo CAR-T therapy targeting CD22 in difficult B-cell cancers. The first patient in its US trial is expected to be treated soon, and initial results from an investigator-led Chinese study are due later this year.
Umoja has also described early safety and activity across several ongoing programs using the same VivoVec delivery platform, although full clinical data have not yet been released. Its solid-tumor candidate UB-VV500 is expected to enter Phase I testing afterward.
A successful in-vivo platform could remove weeks of manufacturing, make repeat treatment easier and bring engineered cells to hospitals without specialist production chains.
Umoja is currently solving delivery before proving it can solve common solid tumors. A2 Bio has the reverse profile: stronger tumor evidence and a much harder manufacturing model.
The eventual winner may combine both ideas, pairing logic-gated targeting with direct in-body cell engineering.
What would prove A2 Bio is really winning?
A2 Bio becomes a real cancer leader only if the next patients turn an exceptional case into a repeatable result.
The first test is response frequency. We need several patients at the same dose to experience meaningful tumor shrinkage, preferably including more complete responses.
The second is durability. Cancer disappearing for a few months is encouraging. Disease staying controlled for several years without further systemic treatment would look much closer to a functional cure.
The third is breadth. Responses should appear in more than one patient and eventually across more than one tumor type. A2 Bio also needs evidence that changing the activating target preserves the same safety advantage.
The fourth is brain control. The isolated central-nervous-system relapse in the first responding patient suggests that engineered cells may have struggled to reach or remain active in the brain. Lung cancer frequently spreads there, so this cannot remain an edge case.
The fifth is manufacturing. A donor-derived product must show that it can survive long enough inside patients to work. Otherwise, A2 Bio will remain dependent on expensive treatments made separately for each person.
The final test is comparison. Early single-arm trials can reveal dramatic activity. A randomized study must eventually show that patients live longer than they would on the best existing treatment.
Until those results arrive, A2 Bio should be treated as the most interesting private contender rather than the likely inventor of a universal cure.
In our biotechnology market deck, we identify pain points entrepreneurs should prioritize
Which startup will solve cancer?
No startup will solve cancer alone, but A2 Biotherapeutics currently has the strongest private therapeutic claim.
A2 Bio leads because its technology addresses a major unsolved problem and has already produced a rare human result in a common solid tumor. ArsenalBio has sophisticated programmable cells without comparable public efficacy data. Affini-T targets an important cancer driver but serves a narrow genetic subgroup. Umoja could transform delivery, though its most relevant solid-tumor evidence lies ahead.
A2 Bio’s lung-cancer result remains one patient inside a 13-person early-stage study. The treatment did not prevent an isolated relapse in the central nervous system, and the broader cohort has yet to produce a clear pattern of responses. A2 Bio’s current therapy is also personalized, slow to manufacture and restricted to tumors with a specific combination of HLA loss and mesothelin expression.
Those weaknesses make failure more likely than success. They do not erase the quality of the underlying result.
Our judgment is direct: A2 Biotherapeutics is the cancer startup to watch today, but calling it the startup that will solve cancer would be premature. The eventual breakthrough will probably combine several technologies, including prevention, early detection, mutation-targeted drugs, programmable immune cells and cheaper in-vivo delivery.
A2 Bio currently owns the most convincing private therapeutic piece of that future. It still needs several more patients to prove that it owns anything larger.
If you want more recent data on this point, please see our latest biotechnology market report.
OUR METHODOLOGY
This analysis tests which private startup currently has the strongest credible claim to materially reduce cancer deaths. We compare therapeutic platforms, early-detection companies and enabling technologies against a demanding standard: evidence in people, relevance to common cancers, repeatability, durability, breadth and the ability to reach large numbers of patients.
We separate an important medical advance from a population-scale solution. A treatment can be transformative for a small genetically defined group without solving cancer broadly, while prevention or early detection can have enormous potential reach without yet proving that they reduce mortality.
Clinical evidence receives the most weight. A confirmed response in one patient can show that a mechanism is biologically possible, but it does not establish a reliable response rate. Randomized survival data, prospective validation and regulatory decisions carry more weight than financing, preclinical results, platform descriptions or management claims.
We also distinguish the quality of a result from the amount of evidence behind it. A2 Biotherapeutics leads because its complete response in metastatic lung cancer is unusually strong for an early solid-tumor cell-therapy program. The conclusion remains provisional because the reported cohort was small, included several tumor types and has not yet produced a repeatable efficacy pattern.
Public companies are used as benchmarks rather than final contenders. GRAIL shows the scale of evidence required in multi-cancer detection, while Revolution Medicines’ randomized daraxonrasib trial shows what convincing therapeutic proof looks like when a platform begins to extend survival in hundreds of patients.
Scalability is treated as part of the medical question, not a separate business issue. A personalized therapy that takes weeks to manufacture and can be delivered only at specialist centers cannot reduce deaths across common cancers unless manufacturing, cost and hospital access improve dramatically.
For prevention and detection, we look for evidence that the intervention changes meaningful outcomes rather than simply finding more abnormalities. Stage shifts, curative surgery rates, emergency presentations and cancer mortality matter more than the number of positive blood tests alone.
We prioritized recent clinical readouts, randomized trials, regulatory decisions, scientific publications and company disclosures with checkable patient numbers and endpoints. Key sources include the International Agency for Research on Cancer on global incidence, mortality and projections, the American Cancer Society on US mortality and survival, and the World Health Organization on preventable cancer burden.
Detection evidence comes primarily from GRAIL’s NHS-Galleri randomized-trial results, ClearNote Health’s Avantect validation data and Harbinger Health’s RESOLVE study.
Therapeutic and platform evidence comes from the National Cancer Institute’s review of CAR-T treatment and manufacturing, the FDA approvals of lifileucel for advanced melanoma and Tecelra for synovial sarcoma, A2 Biotherapeutics’ EVEREST-2 clinical update, ArsenalBio’s clinical-program disclosures and Umoja Biopharma’s in-vivo CAR-T regulatory update.
The final conclusion is not produced by a mechanical score. We assess each contender point by point, give the greatest weight to demonstrated patient outcomes and then judge how much of that evidence could plausibly extend across tumor types and into routine care.
This chart, featured in our biotechnology market deck, breaks down revenue across Europe, Asia, North America, Africa, and South America in the biotechnology market
Related blog posts
- Can cellular reprogramming cause cancer?
- Which startup is setting the pace in AI skin cancer detection?
