Here it is!

This post is sponsored by Consensus, the AI agent for research that now has >2.5 million monthly users. If you’re a doctor, try Medical Mode to query guidelines and the top medical journals. If you’re doing research, Consensus has deals with 6 top publishers, to access full texts that’re behind paywalls. You can use my link for a free trial!

Here’s the transcript from the episode:

[1:39] Tiff: How was ASCO? You went to the biggest cancer conference of the year. It happens annually and is run by the American Society of Clinical Oncology. It was your first ASCO. How was it?

[1:53] Sam: I loved it. I love science, so if you are like me, it is like crack cocaine. Everywhere you go, there is science: posters, talks, and Big Pharma. Eli Lilly is right in the middle, and you can just go and talk to them. It is the latest groundbreaking science, presented for the first time by the researchers themselves. At the posters, you can speak directly to the person who did the work. That person might be the chief executive officer, an academic, or someone else entirely. Chicago is beautiful, by the way.

[2:34] Tiff: That is amazing. Did you go to the Bean?

[2:43] Sam: The mirror thing?

[2:44] Tiff: Yes, the Bean.

[2:45] Sam: Is that what it is called? I should have known that.

[2:47] Tiff: No, it is fine. I think it was being renovated until recently, so you went at a good time.

[2:53] Sam: Yes. You have to go really early, otherwise there are a lot of people and a huge queue.

[2:59] Tiff: In this episode, we thought it would be good to go over some of the highlights from ASCO: the biggest groundbreaking studies, China, and future directions for these therapies.

[3:23] Sam: Exactly. To kick off with the themes of the conference, I thought there were five major themes at ASCO this year. The first was pancreatic cancer. The major news was daraxonrasib. I can say that now that it has been talked about so much.

[3:48] Tiff: Daraxonrasib. I love the name, because “on” refers to the on state, “RAS” is the target, and “-ib” means it is a small-molecule drug.

[4:00] Sam: Exactly. Daraxonrasib was the major news. It increased overall survival in second-line metastatic pancreatic cancer compared with chemotherapy. It also improved quality of life, reduced pain, and did this with almost half as many serious adverse events.

[4:19] Tiff: It got a standing ovation from the crowd. I saw your tweet. It went viral.

[4:26] Sam: Super viral. Very exciting. I loved it. That was the main news, but pancreatic cancer had more than one breakthrough. Daraxonrasib stole the show, but there were many other therapies: antibody-drug conjugates, and a really nice bispecific T-cell engager. There is an antigen that pancreatic cancer sometimes expresses called claudin 18.2.

[4:55] Tiff: Very specific.

[4:57] Sam: Yes. It is also expressed in some gastric cancer cells. If you target claudin 18.2 and CD3 to recruit T cells, and give that to pancreatic cancer patients, that also looks like quite an effective approach.

[5:10] Tiff: That is amazing. There was not only one breakthrough in pancreatic cancer. There were several. For years, even decades, there have only been incremental improvements in pancreatic cancer, which has a very poor five-year survival rate of around 10%. Suddenly, there has been a lot of progress announced within one year. I thought that was really interesting.

[5:45] Sam: It is. Revolution Medicines is the company behind daraxonrasib, but now there are many companies with RAS inhibitors, and they are approaching it from every direction. Big Pharma is there, small biotech is there. They are developing pan-RAS inhibitors, KRAS G12C inhibitors, KRAS G12D inhibitors, on-state inhibitors, off-state inhibitors, and different combinations.

[6:08] Sam: This week, there was an impressive combination of daraxonrasib plus another new drug in pancreatic cancer, and it improved survival even further on top of what daraxonrasib does.

[6:23] Tiff: That is incredible.

[6:24] Sam: What I think we are going to see is daraxonrasib becoming the foundation, or bottom layer, of therapy, and then being combined with other therapies.

[6:32] Tiff: For that specific mutation or subgroup.

[6:36] Sam: Exactly. I did not think we would see this layering of therapy so quickly, given that it has only been two weeks since we saw the data in the first place. But this week we are already seeing it start. The company I am talking about is Tango Therapeutics. They are working with Revolution Medicines to build this combination therapy.

[6:56] Tiff: That is so exciting. They waste no time.

[6:59] Sam: They do not. For those who missed it, daraxonrasib doubled survival in PDAC, which is the most common form of pancreatic cancer, with a hazard ratio of 0.4. It decreased pain and side effects, and it is a daily oral pill.

[7:31] Tiff: That is important. People talk about the time burden of cancer therapy. If you are having chemotherapy as an intravenous treatment, you have to go to the clinic.

[7:44] Sam: Exactly. You have to go to the hospital, sit there, and have the intravenous chemotherapy treatment. With daraxonrasib, it is a pill. You take it at home. It reduces the time burden and time toxicity, as well as other toxicity. It does not mean it is completely without toxicities. There were rashes, stomatitis, and one patient in the trial died because of lung inflammation. It is not completely without adverse events, but it is cleaner than chemotherapy.

[8:14] Tiff: That is huge. It is also a cool paradigm shift from the last generation of RAS inhibitors, which tried to target the off state of RAS. RAS is an oncogenic driver of cell division. It is like a gas pedal that is being pressed down all the time, and we have to find a way to stop it.

[8:40] Tiff: The issue with RAS, as we covered in a previous episode before the phase 3 data came out, is that it was considered an undruggable target because there is no pocket in the molecule for a drug to bind to. Revolution Medicines became creative and developed a molecular glue that binds to the chaperone protein cyclophilin A. That creates a complex that can target the on state of RAS rather than the off state. The on state is more susceptible to inhibition, because most of the time RAS is on in human cancer cells rather than off. So you might as well target the on state.

[9:36] Sam: Exactly. If you target the off state, your drug is waiting around for RAS to switch itself off, bind, and then inhibit. But the RAS protein spends hardly any time in the off state. It spends most of its time in the on state, so ideally you target on, not off.

[9:55] Tiff: It is clever. The craziest thing is that this is a pan-RAS inhibitor. In oncology, KRAS is the most common mutated form of RAS, but there are also NRAS and HRAS. Even within KRAS, there are different mutations. There have been developments before in G12C mutations, which means the glycine at codon 12 is replaced by cysteine. In pancreatic cancer, people care a lot about G12D.

[10:37] Sam: Yes, where it is replaced by aspartic acid.

[10:40] Tiff: Now they do not have to worry as much about which type of mutation is occurring in the tumour.

[10:46] Sam: Exactly. That is also one of the future directions. RAS is mutated in pancreatic cancer, but also in colorectal cancer, lung cancer, and other cancers. The big three are pancreas, colorectal, and lung.

[11:02] Tiff: Lung is what Revolution Medicines is going after next in one of their trials.

[11:06] Sam: Exactly. It may not only revolutionise the treatment of pancreatic cancer, but also lung, colorectal, and other cancers. One future direction in cancer is RAS inhibition as a therapeutic foundation.

[11:22] Tiff: That is exciting. Revolution Medicines is a Californian company founded in 2014, and they IPO’d in 2020. Did you look at the stock price afterwards?

[11:34] Sam: Yes. I am in the stock. I think most biotech people are in the stock. Why would they not be?

[11:45] Sam: The market cap is now around $35 billion. People start to think about acquisition: is this going to get bought by Big Pharma? But if the market cap is $35 billion, Big Pharma has to pay a premium, so they would have to pay $40 billion or $50 billion.

[11:59] Tiff: They use the current stock price as a benchmark.

[12:03] Sam: Exactly. As Big Pharma, how do you spend $35 billion, $40 billion, or $50 billion on one company? That probably means it is too big to get taken out now, which is interesting.

[12:15] Tiff: The biggest deals we have seen in the past year have been around $10 billion.

[12:20] Sam: This is much bigger. Some companies could technically take it out, but it would be a huge swing.

[12:27] Tiff: I can see smaller deals being made. Maybe not the whole company, but licensing one asset or one indication to another Big Pharma company.

[12:38] Sam: Exactly. There is another company called Erasca with a molecule that is basically identical to Revolution Medicines’ daraxonrasib. If you look at the chemical structures, they are almost the same, so there is a patent concern. If you are Big Pharma and want to buy something similar more cheaply, maybe you take out Erasca.

[13:02] Tiff: Then they have to deal with the patent disputes, not Big Pharma.

[13:09] Sam: That was the first theme: progress in pancreatic cancer.

[13:15] Tiff: The second theme was other innovative therapies. What else did you see?

[13:21] Sam: Daraxonrasib was definitely the most important one, but there were lots of others. One was ivonescimab, a bispecific antibody.

[13:29] Tiff: PD-1 and VEGF.

[13:30] Sam: Exactly. It is a PD-1/VEGF bispecific made in China, with trials done in China. In lung cancer, it beat PD-1 immunotherapy on its own. PD-1 immunotherapy is one of the biggest drug classes in the world, but if you use a bispecific that also blocks VEGF, it improved overall survival in lung cancer.

[13:53] Tiff: That is incredible. It is good synergy. PD-1 is the checkpoint inhibitor part of the drug, and the VEGF part is supposed to promote a more normalised vasculature. Then immune cells, especially T cells, can reach the tumour and kill tumour cells. When VEGF is overexpressed by cancers, the chaotic vasculature can prevent immune cells from getting to the tumour.

[14:25] Sam: Exactly. The trial format at ASCO is fascinating. In the plenary session, the lead investigator presents the trial, and then someone else puts the trial in context and criticises it. So the lead investigator is sitting there while someone else says, “Yes, but...”

[14:49] Tiff: It is like a lab meeting in front of thousands of people.

[14:53] Sam: Exactly. The main criticism was that all of the data were from China. The data are strong, but from that perspective, it only really applies to the Chinese population. If we want to show that ivonescimab can beat Keytruda, or pembrolizumab, in a broader population, we have to do the trial in the US. That is happening.

[15:22] Tiff: It has to be tested across different ethnicities. The same statement can be reversed as well. American therapies that want to enter the Chinese market, which is obviously huge, also have to do trials in China. That is happening too. It goes both ways.

[15:42] Sam: It does. But it was interesting that this was the main criticism. Ivonescimab is definitely a major breakthrough. We might actually be able to beat PD-1.

[15:55] Tiff: I hope they are working with MD Anderson or another major centre to test it in a Western population.

[16:00] Sam: They are running at least 15 massive global trials with that compound.

[16:08] Tiff: Amazing.

[16:09] Sam: Another innovative therapy I liked was for cancer cachexia. We talked about this in our first pilot episode. This is one of your favourite proteins: GDF-15.

[16:22] Tiff: Exactly.

[16:23] Sam: Cancer cachexia is a wasting syndrome in advanced cancer. Patients lose muscle and weight, lose appetite, and feel miserable. There are no good drugs. It seems to be driven by a stress-induced cytokine called GDF-15.

[16:42] Sam: This new study, also from China, tested a GDF-15 antibody as a bispecific with an IL-6 blocking antibody. It reduces inflammation and blocks the cachexia-driving factor GDF-15. I spoke to the person presenting the poster, and they told me that the main complaint from patients in the trial was that they were gaining too much weight.

[17:08] Tiff: That sounds like a champagne problem.

[17:14] Sam: With this bispecific antibody, they were gaining too much weight. Some gained around 25% of their body weight.

[17:20] Tiff: That is amazing.

[17:23] Tiff: That is such a cool synergy: block GDF-15 and block IL-6. What does blocking IL-6 do here? Does it just reduce inflammation? Is that because blocking GDF-15 might increase inflammation?

[17:46] Sam: I think one of the mechanisms by which GDF-15 is induced is probably systemic inflammation.

[17:53] Tiff: So you block GDF-15 directly, but also suppress its induction by suppressing systemic inflammation. To be honest, I do not know. It is just cool. I wonder how they make these decisions about what to pair in a bispecific.

[18:11] Sam: As we have discussed many times, suppressing chronic systemic inflammation is generally a good idea. People with advanced cancer have high inflammation. If you suppress it, you probably make them feel better. They also say it increases appetite, increases muscle, increases weight, and improves quality of life. It is generally good to control inflammation.

[18:32] Tiff: We could have a whole other episode on inflammation. I know it is one of your theses.

[18:40] Sam: It is definitely a thesis.

[18:43] Tiff: What else was there apart from this cachexia drug?

[18:48] Sam: Antibody-drug conjugates were everywhere. They are so hot now. The biggest one is sac-TMT. I think they are running 17 phase 3 trials with sac-TMT.

[19:05] Tiff: It is American now. It is Merck. But it was produced in China.

[19:10] Sam: It was developed in China and acquired by Merck. They are running 17 phase 3 trials. It is a TROP2-targeted antibody-drug conjugate with a topoisomerase I inhibitor payload.

[19:22] Tiff: So it is an antibody-drug conjugate delivering a topoisomerase I inhibitor payload.

[19:31] Sam: Exactly. It also improved overall survival in lung cancer compared with standard of care.

[19:40] Sam: If you add daraxonrasib plus those three innovative therapies, that is the second major theme. The third theme was prevention. I was at ASCO trying to discover what was happening in prevention, because prevention is the best cure.

[20:00] Tiff: Or early intervention.

[20:05] Sam: Exactly. There was a lot on GLP-1s: many observational studies showing that GLP-1s seem to reduce the risk of many different cancer types.

[20:15] Tiff: That is cool.

[20:16] Sam: Basically everything you can think of: breast cancer, colorectal cancer, blood cancers, and lung cancer.

[20:22] Tiff: Somehow, it does not come as a surprise that they could have those effects.

[20:26] Sam: Obesity is a risk factor for cancer, so GLP-1s obviously reduce that risk factor. But there must be something else going on as well.

[20:37] Tiff: Maybe inflammation. I do not know the effects of GLP-1s on inflammation.

[20:44] Sam: Systemic inflammation definitely drives cancer, so suppressing inflammation is likely to be one mechanism.

[20:55] Tiff: So far, these are observational studies, because there is now so much data on GLP-1s and researchers can correlate them with other datasets. The next step is randomised trials to support this.

[21:12] Sam: What is amazing about the conference is that Eli Lilly’s oncology team is just standing there. Lilly makes tirzepatide, so you can go and ask them what they think. I asked whether they were going to run randomised trials with GLP-1s to prevent cancer. They said no, but if you look deeper, Lilly actually is running relevant trials.

[21:39] Sam: There is a randomised trial with tirzepatide where liver cancer is at least a readout. There is another trial with retatrutide and tirzepatide, the dual and triple agonists, looking at whether all-cause cancer risk is reduced. Cancer is not the primary endpoint, but it is an exploratory endpoint.

[22:03] Tiff: There is another trial happening in breast cancer for primary and secondary prevention with a GLP-1. These are already underway.

[22:13] Sam: Yes. We will have readouts in a few years.

[22:17] Tiff: It feels like everyone is going to be on a GLP-1 soon, for one reason or another.

[22:25] Sam: What is fascinating is that everyone knew overweight and obesity were bad, but GLP-1s have shown us they are worse than we thought. Obesity contributes to so many diseases, and when you suppress it, many things improve.

[22:42] Tiff: It feels like a holy grail, with obvious caveats and side effects.

[22:52] Sam: Do not worry about side effects.

[22:53] Tiff: Pancreatitis, which you disproved.

[22:55] Sam: Yes. No pancreatitis, actually. That was the third theme. The fourth theme was cancer vaccines.

[23:03] Tiff: One of my favourites. The big Moderna and Merck vaccines are personalised vaccines in phase 2. The phase 2 trial took five years to finish. It was in resected melanoma, using the vaccine plus pembrolizumab. It was around 72% versus 49% progression-free survival.

[23:50] Sam: It is definitely bullish. There is also a Chinese company running a trial in oesophageal squamous cell cancer with a personalised vaccine. Oesophageal cancer has miserable survival, and they got strong results. I was happy because Infinitopes is running a trial too.

[24:18] Tiff: A different kind of trial. You are an investor, obviously. We are doing ours in oesophageal adenocarcinoma. Oesophageal adenocarcinoma and oesophageal squamous cell carcinoma overlap, but they have distinct profiles. Squamous disease mainly affects Asian and Chinese populations, while adenocarcinoma mainly affects Western populations.

[24:45] Sam: That is a good point. It is useful that they are running the squamous trial in China, where it is more relevant. It would be interesting to see whether any antigens overlap between the two types. Their trial is personalised, whereas Infinitopes is doing semi-personalised, off-the-shelf vaccines. It is a different approach, but as proof of concept showing that you can improve survival in a difficult cancer with a vaccine, it is very bullish for Infinitopes. That is why I was happy to make that discovery.

[25:17] Tiff: Watch his face, everyone.

[25:21] Sam: That was the fourth theme. The fifth theme is that, with the exception of daraxonrasib, a lot of the most innovative work we discussed came out of China.

[25:35] Tiff: China biotech is on the stage.

[25:40] Sam: China biotech is doing amazing work.

[25:45] Tiff: I was surprised. China is no longer just doing me-too therapies. They are innovating and coming up with their own ideas. You said you had some stats to back that up.

[26:04] Sam: Some criticism of China is that it is doing me-too therapies and fast follows. That is happening, and China is also the pharmacy of the world. But they are also doing innovative things: ivonescimab, sac-TMT, and the cachexia bispecific we discussed.

[26:21] Sam: Last year, there were 5,000 new trials started worldwide for innovative therapies. Half of them started in China. In 2015, it was about 10%. There were about 2,000 trials started globally and 200 in China. In 2025, it was 5,000 globally and 2,500 in China. China went from around 10% of innovative trial starts to 50%.

[26:53] Sam: Not only is the number of innovative trials increasing rapidly, showing that science is speeding up, but China is also playing a bigger and bigger role.

[27:07] Tiff: How many of these do you think are sponsored by American or European Big Pharma companies? A lot of them partner with Chinese pharma because they get access to the Chinese population, which is a major market.

[27:27] Sam: That is a good question. This year so far, for the first time ever, most biotech deal value is going to China. Around 70% of biotech deal value is going to Chinese companies. In 2021, it was 4%.

[27:50] Tiff: That is huge.

[27:52] Sam: It has gone from 4% to 70%, stepping up over time.

[27:59] Tiff: Why do you think that is? Is it because the companies are cost-effective, the trials are cheaper, and the trials are faster?

[28:07] Sam: Exactly. It is cheaper and faster. The cost per patient in a trial in the US is around $65,000. In China, it is around $20,000 to $25,000. It is also much faster to recruit patients, about three times as fast. It is roughly a third of the cost and much faster.

[28:30] Tiff: Those are basic business fundamentals. What is faster and cheaper gets you to the end goal faster.

[28:39] Sam: Exactly. If we want to reverse the trend of all this money going to Chinese companies, we have to fix the fundamentals in the US and Europe. I think that is doable. We need to lean more on artificial intelligence.

[28:55] Tiff: We also need to look at the architecture of the regulatory bodies. Part of the speed of running trials in China may come from the NMPA enabling cross-border trials to happen quickly. I wonder whether it gets criticised as being less stringent than the FDA. Is that why trials can be run so fast? Is that why people are going there? Is the data reliable? These are questions some founders have asked. As a founder myself, I do not see an obvious reason to stop me from going to China and running a trial there.

[29:56] Sam: To prove the concept quickly, China is an amazing place to do it. But to get your product approved in Europe, the US, or both, you still have to run the trial in those regions as well. You still need to show that the results apply to the relevant broader population.

[30:20] Tiff: Now it feels like China is becoming a proof-of-concept ground for proving therapies quickly. Once the proof of concept has been shown in China, companies can bring the therapy back to the West. It swings both ways and uses the best of both worlds.

[30:44] Sam: Exactly. In general, as I said on TBPN this week, if you are on the side of new medicines and patients, more competition is better. Competition is good for patients. If China is just as good as the US, or arguably even better, at doing this, that is great. We will get more medicines faster, which is better for patients, saves more lives, improves lives, and helps eradicate disease.

[31:17] Tiff: That is the end goal. One good thing about medicine and science is that they are cross-border. As you saw at ASCO, researchers fly in from many different countries. That is a beautiful thing about science.

[31:36] Sam: It is like music. It is not contained within one country.

[31:43] Sam: That was the China theme. I also wanted to say something about the future, given what we have discussed.

[31:52] Tiff: I am excited.

[31:56] Sam: These are basically syntheses of what we have already said. First, RAS inhibition is going to become a foundation of therapy in several cancer types: pancreatic cancer, colorectal cancer, and lung cancer.

[32:19] Sam: Second, we are going to see many more antibody-drug conjugates. When I spoke to some of the pharma teams and asked what the most exciting product in their pipeline was, the Lilly team told me about a new antibody-drug conjugate. It does not only target topoisomerase I, like many antibody-drug conjugates do, but also targets ATR.

[32:44] Sam: That takes me back to our PhD days. It inhibits both ATR and topoisomerase I. If you do that, you disrupt DNA replication, which is disastrous for the cell and can kill it.

[32:57] Tiff: That makes sense. Remember the problem we had in our PhDs? This is a niche reference. ATR is not essential for the cell to live.

[33:10] Sam: Is that right? Do you mean in terms of inhibition?

[33:14] Tiff: No, I am confusing it with ATM. ATM is not essential. My morning coffee has not kicked in. Does targeting both make it more specific?

[33:30] Sam: I think so. To be honest, the product is preclinical. They acquired it. If you inhibit topoisomerase I, ATR helps repair the topoisomerase I lesions.

[33:44] Tiff: So they are synthetic lethal.

[33:48] Sam: Exactly. It is a synthetic lethal or synergistic relationship.

[33:53] Sam: The second future theme is many more antibody-drug conjugates. The third is radiopharmaceuticals. When I asked what they were most excited about in the pipeline, they also mentioned radioisotopes.

[34:07] Tiff: Targeted radiopharmaceuticals.

[34:10] Sam: Exactly. It is similar to an antibody-drug conjugate, but instead of using chemotherapy, you use a radioisotope, such as an alpha or beta particle. You link it to a tumour antigen and use it to kill cancer cells.

[34:25] Tiff: That is a creative payload. There have been a few more biology-physics convergence therapies coming out.

[34:34] Sam: The third future theme is radiopharma. The fourth is prevention.

[34:41] Tiff: That is your job. I am working on it.

[34:46] Tiff: Thank you for being an investor in both companies.

[34:54] Sam: The fifth theme is that daraxonrasib has shown that things we thought were undruggable are not necessarily undruggable. There will now be a wave of supposedly undruggable things being drugged. My two top examples are MYC and p53. Both have companies focused on drugging them, and I think we will see them drugged.

[35:24] Tiff: I completely agree. It is about being creative. You do not have to drug the target directly. You can drug its interaction partners.

[35:38] Sam: Exactly. That is one way people are trying to target MYC. Another is degraders, such as PROTACs. You do not need a little small molecule. You can use a PROTAC.

[35:51] Tiff: We now have much more context around disease biology. With RAS, we now know to go for the on state and not the off state. The same principle will apply to MYC, p53, and other previously undruggable targets.

[36:10] Sam: I am really excited.

[36:13] Sam: Whenever we talk about therapies, we also have to say that prevention is best. Early detection and prevention are much better than treating metastatic disease. We have to combine these approaches.

[36:22] Tiff: That is why we have early detection technologies. They make existing therapies work better.

[36:30] Sam: Exactly. You have to jump in before the disease becomes metastatic.

[36:38] Tiff: You are working on that, and we will keep people updated on the progress.

[37:59] Sam: One last thing. There was an interesting All-In Podcast episode this week. They had someone from EcoR1 Capital, which is a biotech hedge fund.

[38:17] Tiff: Very nerdy name.

[38:21] Sam: He gave a stock pitch for a company called Aktis Oncology. Do you know what they work on? Radiopharmaceuticals. Radiopharma is definitely one of the major up-and-coming themes.

[38:37] Tiff: Now that cancer vaccines have matured, the next new shiny thing is radiopharmaceuticals.

[38:46] Sam: There are already two approved radiopharmaceuticals, including one in prostate cancer. But there are only two, so it is still very early. It is definitely a direction for the future.

[38:57] Tiff: Exciting. I am glad you had a good time. Let’s go next year.

[39:01] Sam: Come next year and we will do some chatting, interviewing, or something like that.

[39:08] Tiff: More standing ovations.

And the Youtube video:

1. The RAS revolution continues

After decades of minimal progress, a couple of weeks ago we saw the first step change in survival for metastatic pancreatic cancer — with the RAS inhibitor, Daraxonrasib.

This week a Boston company, Tango Therapeutics, revealed new data that layers a new therapy on top of Daraxonrasib. Their approach depends on a concept called synthetic lethality.

Tango’s approach is rooted in a couple of discoveries from 2016 (the first and the second). About 40% of pancreatic cancers are deficient in the metabolic enzyme MTAP, which clears a metabolite called methylthioadenosine (MTA). This means MTA accumulates, partly suppressing an epigenetic enzyme, PRMT5, and making tumor cells hypersensitive to pharmacological PRMT5 inhibition (with Vopimetostat). PRMT5 is an essential gene; without it, cancer cells die.

I am comparing cross-trial here — and this is only a phase 1/2 trial — but the numbers look good. In MTAP-deficient pancreatic cancer, the objective response rate and 6-month progression-free survival were 92% and 90%, respectively, with the Vopimetostat-Daraxonrasib combination, compared with 20-32% and 41-56% with Daraxonrasib alone (and 11% and 33% with the current standard-of-care chemotherapy).

The synthetic lethality itself is independent of RAS, but it looks like PRMT5 inhibition (in MTAP-deficient cancers) might also be synergistic with RAS inhibition.

This layering of therapy is likely to continue, and could continue to improve outcomes, with RAS inhibition as the foundation. Since most pancreatic cancer still presents as metastatic, what we need alongside this is better early detection.

2. A potentially best-in-class ulcerative colitis drug

Last July, a small French biotech company dropped some data that led to the biggest stock price jump in Nasdaq history. That company is Abivax, who have an experimental new drug for inflammatory bowel disease.

Trials in inflammatory bowel disease happen in two stages: first, induction of remission, and then maintenance of remission. Last July’s data were for induction, and this week (actually it was last week, I am a bit late) Abivax dropped new maintenance data.

Their drug is Obefazimod, a daily oral pill, and the indication here is ulcerative colitis (though they’re also running a trial in Crohn’s). The mode-of-action is a bit up in the air, but it seems to work by inducing expression of miRNA-124, an anti-inflammatory miRNA. The way Obefazimod probably does this is through modulation of the aryl hydrocarbon (AH) receptor. Strangely enough, this was discovered not by Abivax, but by another company with a similar drug, Equillium.

Here are the data, looking at clinical remission (stool-frequency score, rectal-bleeding score, and Mayo endoscopic subscore). Efficacy looks like it might be best in class — better even (cross-trial) than the injectable biologics:

Endoscopic remission is a more objective endpoint, which measures remission on colonoscopy. Here, the data are arguably even better:

Clearly safety is essential here. The safety endpoint I’m most interested in is opportunistic infections, which can happen with some of the other ulcerative colitis drugs, given their immunosuppressive nature. With Obefazimod, the risk of opportunistic infections was 1% with placebo, 1% with low-dose Obefazimod, and 0.5% with high-dose Obefazimod: no clear concerns — the drug is not immunosuppressive.

Listening to gastroenterologists discuss this, Obefazimod has potential first-line use as well as later-line use — indeed, patients in this trial had typically failed 1, 2, 3, or even 4 prior therapy classes before Obefazimod.

The next landmarks here are submission to the FDA this year, and the Crohn’s phase 2 trial next year.

3. Discovery of a defined subtype of inflammatory bowel disease

In inflammatory bowel disease (Crohn’s disease and ulcerative colitis) there is a small amount of therapy tailoring, but it’s really limited to a few things — disease severity, prior biologic exposure, drug safety profile. Choice of therapy itself is pretty much based on trial and error: try one drug, and move to the next if it doesn’t work.

This is partly because the causes of inflammatory bowel disease are still poorly defined, and no established, common molecular mechanisms exist. This study, from the New England Journal this week, changes that.

Read more

1. A standing ovation for Daraxonrasib

The story of the meeting was Revolution Medicines’ full phase 3 dataset, for Daraxonrasib vs. chemotherapy in metastatic pancreatic cancer. This was greeted by a standing ovation (in the middle of the talk!) which I recorded here:

Here’s the full Kaplan-Meier curve:

There are two reasons this is so important. First, many trials have been done in this setting, all of which failed to improve (or only incrementally improved) the standard of care. Daraxonrasib inhibits the oncogenic driver in pancreatic cancer, RAS, producing this step change in survival and laying the new foundation of therapy:

Second, RAS was king of the undruggable, and it took some serious creativity to inhibit it. Daraxonrasib is a 'molecular glue' that locks RAS into an inactive trimer with a third molecule, Cyclophilin A. This works for many species of RAS — both wildtype and mutant forms — and Daraxonrasib binds RAS in its ON state (which produces more potent inhibition than OFF state inhibition). This also gives rise to the name: Darax-ON-RAS-IB.

As well as doubling average survival, Daraxonrasib improved quality of life and reduced pain associated with pancreatic cancer:

What's next for Daraxonrasib? More trials! 1st line in metastatic pancreatic cancer, and after surgery in resectable disease. RAS is also a major driver in lung cancer and in other solid cancers, like colorectal, so Revolution Medicines are running trials there, too:

Daraxonrasib is far from the only RAS inhibitor moving through trials — these are some of the other ongoing phase 3 trials. There are many others in earlier phases, too – the more the merrier!

2. Innovative therapies

There were trials presented for new, innovative therapies — antibody-drug conjugates (ADCs), vaccines, bispecific and even trispecific(!) antibodies.

A bispecific antibody in cancer cachexia

One that caught my eye was this trial in cancer cachexia (weight and muscle loss in advanced cancer). There are no targeted therapies approved for this condition, but it’s become pretty well-established that a stress-induced cytokine, GDF15, drives it, and recently were recently positive phase 2 data on a GDF15-blocker from Pfizer.

This approach from ASCO builds on that idea with a bispecific antibody that inhibits both GDF15 and IL6 — blocking the driver of cachexia (anti-GDF15) and suppressing inflammation (anti-IL6).

In patients with advanced cancer and cachexia, the antibody increased bodyweight (up to 25% for high dose vs. low dose), increased muscle mass, and increased appetite. The presenter of the poster told me that patients’ main complaint was... gaining too much weight.

Cancer vaccines

These new, final, 5-year phase 2b data were presented for Moderna/Merck’s personalised cancer vaccine (plus the standard of care, Pembrolizumab, a PD1-blocking antibody) in resected melanoma. Melanomas were removed by surgery, personalised vaccines built based on specific cancers’ mutations, with the aim to prevent recurrence — and it worked: there was less recurrence and less metastasis.

Based on a similar idea, these data, from a Chinese biotech, dropped in esophageal squamous cell carcinoma. This condition has huge unmet need, with average disease-free survival of about 22.4 months with the best standard-of-care (PD1 blockade). This personalised neoantigen vaccine looks like it might extend disease-free survival, when used after surgery:

Beyond this, there were new data on a trispecific anti-PD1/CTLA4/VEGF antibody in lung cancer, new data on CEACAM5-targeted CAR-T cells in colorectal cancer, and data on a neoantigen vaccine in glioblastoma, where there’s been minimal progress for decades.

3. Pancreatic cancer progress beyond Daraxonrasib

Read more

1. A new class of blood pressure drug

Hypertension is not a trendy part of medicine, but it’s common and important — high blood pressure increases the risk of cardiovascular disease (heart attacks and strokes). In general, lower tends to be better, with a sweet spot for systolic blood pressure around 110-120 mm Hg:

There are lots of blood pressure drugs that target hypertension, from many different angles, but treatment-resistant hypertension — hypertension despite treatment with ≥ 3 drugs — still affects ~7-10 million adults in the US.

Aldosterone signalling is an established hypertension target, and drugs that block the aldosterone receptor (like spironolactone) are the go-to medicines for resistant hypertension. Inhibition of the receptor can actually increase aldosterone levels though, via negative feedback, which causes its own problems.

For these reasons, AstraZeneca tried a different approach: blocking the synthesis of aldosterone altogether. Their drug is Baxdrostat (-stat because it’s an enzyme blocker and -dro because its target is aldosterone synthase) and was just approved this week.

Tested in two trials (the first and the second) in patients taking an average of 4 blood pressure meds at baseline (a diuretic, an ACEi or ARB, a calcium channel and a beta blocker — mineralocorticoid receptor antagonists were not allowed!) Baxdrostat nicely reduces blood pressure. Patients here wore continuous blood pressure monitors, so you can also see how blood pressure fluctuates throughout the day — higher in the day, lower at night:

Today, the hypertension market is worth about $25B, and AstraZeneca forecast $5B peak sales for this drug. They are the first to this class but they will not be the last: Mineralys Therapeutics and Boehringer Ingelheim both have similar, late-stage drugs.

2. The first phase 3 trial for Retatrutide in obesity

Retatrutide is Eli Lilly’s ‘Triple G’ — a triple agonist of the GLP1, GIP, and glucagon receptors. It is not yet approved but it’s already used among biohackers and bodybuilders. This week even Elon Musk was talking about it.

Mechanistically, GLP1 reduces hunger and slows gastric emptying; GIP stimulates further weight loss and adds a tolerising factor that reduces side-effects; glucagon agonism helps to burn fat.

Retatrutide already has positive trials in osteoarthritis and fatty liver disease, and, this week, the first phase 3 data in obesity dropped.

Read more

Read more

1. ctDNA helps detect cancer and guide therapy

Circulating tumor DNA (ctDNA) — DNA released from tumors, shed into blood — has a few different uses in cancer: from monitoring response to therapy to identifying targetable mutations, to predicting relapse and population screening.

This week, the first ctDNA-based approach was approved to guide cancer therapy based on molecular residual disease — in muscle-invasive bladder cancer, where around 50% of patients have recurrence even after successful surgery (cystectomy, removal of the bladder).

Current surveillance largely relies on imaging and clinical features like symptoms to detect recurrence — ctDNA can detect molecular evidence much sooner.

The approval is based on this trial, where molecular residual disease was detected using Natera’s Signatera diagnostic kit (also approved this week, to go along with the therapy). Those with residual disease, detected by positive ctDNA, who received the PD-L1 inhibitor, Atezolizumab, had better survival than those who received placebo:

This might also mean that those with negative ctDNA can avoid unnecessary therapy.

The excitement here is real — and reflected in the Natera stock price over the last few years:

There are many other trials using the same approach to detect recurrence and guide treatment decisions — for example, in colorectal cancer, breast cancer, and lymphoma.

Given the sensitivity of cancer detection using ctDNA, there is even hope that this approach could be used to help patients avoid surgery after neoadjuvant therapy — there’s evidence that might work in bladder cancer, rectal cancer, and breast cancer.

2. A Tau depleter might slow Alzheimer’s (but misses the primary endpoint)

Tau is central to the Alzheimer’s cascade — it aggregates as oligomers and fibrils, degenerating neurons in the hippocampus. This starts many years before diagnosis, and leads to loss of hippocampal volume and cognitive decline:

Diranersen is an antisense oligonucleotide, developed as a collaboration between Ionis and Biogen, that depletes the mRNA encoding Tau. In a previous phase 1 trial, it effectively prevented accumulation (and even helped clear) Tau from the brains of patients with Alzheimer’s:

This week, the phase 2 trial read out in patients with early, symptomatic Alzheimer’s. The primary endpoint was dose-responsiveness of Diranersen on cognitive decline (measured by Clinical Dementia Rating–Sum of Boxes, CDR-SB). We don’t have the detailed data yet, just the headline: cognitive decline was slowed, but this was not dose-responsive (the effect wasn’t stronger when the dose of Diranersen was increased).

So this is a mixed bag — the lack of dose responsiveness could indicate statistical noise rather than genuine therapeutic effect — though this would certainly not be the first time a drug ‘fails’ in phase 2 and succeeds in a subsequent phase 3. Biogen have said they will now progress this to phase 3, so either way, we will find out!

Just to highlight the heat of Tau as a target, Biogen/Ionis are not the only ones working on this — there is a lot of attention here:

3. A pill for Covid prophylaxis

This week, we saw results for the very first oral pill proven to prevent Covid after an infectious exposure. This has been a long time coming — here’s a short history of the trials that have been done on this:

1. Ensitrelvir (2026) — Succeeded

2. Casirivimab/imdevimab (2021) — Succeeded historically, but variant-dependent

3. Bamlanivimab (2021) — Succeeded historically, but variant-dependent

4. Tixagevimab/cilgavimab (2022) — Failed

5. Paxlovid (Nirmatrelvir/ritonavir) (2024) — Failed

6. Molnupiravir (2023) — Failed

7. Hydroxychloroquine (2021) — Failed

8. Lopinavir/ritonavir (2021) — Failed

9. Favipiravir (2026) — Failed

10. Ivermectin (2020) — Failed

Here are the data — household contacts of people with Covid (within 3 days of the index patient’s symptom onset) were given Ensitrelvir or placebo for 5 days:

On the basis of these data, Ensitrelvir was approved in Japan in March — and other regulators around the world are looking at it now too. It could be particularly useful for occupational exposure in hospitals, or for people who are very high risk (for example, in care homes).

Read more

1. A breakthrough cardiac drug

The most common genetic heart condition is hypertrophic cardiomyopathy — it affects about 1:200-500 people (though not all of these manifest symptoms). It causes thickening of the left side of the heart, due to a range of mutations — most commonly, in the proteins that help heart muscle contraction, like myosin.

This hypertrophy can obstruct the ejection of blood from the heart (via the left ventricular outflow tract), which happens in 50-70% of diagnosed patients:

There are two drugs approved to help here — Bristol Myers Squibbs’ Mavacamten and Cytokinetics’ Aficamten. They work in the same way: inhibition of cardiac myosin, which means the heart muscle contracts less forcefully and the obstruction is relieved.

The major unmet need remains the 30-50% of patients with non-obstructive disease, where there are no disease-modifying therapies, but there’s still a risk of heart failure, arrhythmias, and sudden cardiac death.

Bristol’s trial with Mavacamten, which read out last year, was negative for this indication (below, left). This week, Cytokinetics read out a similar trial with Aficamten, which was resoundingly positive (below, right):

The primary outcome here is peak VO₂, which is basically a measure of exercise capacity. The other readout is ‘KCCQ-CSS’, which measures symptom burden and physical limitation — Aficamten was positive there too (Mavacamten was negative).

If approved, Aficamten will become the first drug specifically approved for non-obstructive hypertrophic cardiomyopathy, which affects about 5–12 million people globally.

2. The first published clinical Daraxonrasib data

This year’s breakthrough of the year is likely to go to Daraxonrasib, which recently read out a positive phase 3 trial in 2nd line metastatic pancreatic cancer, and a positive phase 1/2 in the 1st line. But this week the first clinical data were published (in the NEJM — where else?).

Daraxonrasib targets RAS, the most commonly-mutated oncogene in cancer — it’s a driver in ~90% of pancreas cancers, ~30% of lung adenocarcinomas, and ~40% of colorectal cancers.

Daraxonrasib is a molecular glue which binds to a scaffolding protein (called Cyclophilin A) and then RAS. This trimer locks RAS into an inactive state that’s unable to activate the downstream pathway, and promotes GTP hydrolysis — to push RAS into the ‘off’ state.

The published data are from a phase 1/2 trial in metastatic pancreas cancer. In the 2nd line of therapy, median overall survival was 15.6 months (which roughly matches the press-released top line from the phase 3) — for context, overall survival with standard-of-care chemotherapy in this setting is typically about 5-7 months.

In the 3rd line of therapy, median overall survival was 9.0 months. Overall survival here, with standard-of-care chemotherapy, is about 5.5 months.

What’s next for Daraxonrasib?

The next milestone will be the full presentation and (probably) co-publication of the full randomized 2nd line data in metastatic pancreatic cancer at the end of May at ASCO. After that, regulatory submissions and 2nd line approval later this year. In 2027, we’ll see completion of the non-small-cell lung cancer trial; in 2028, completion of the first-line metastatic pancreatic cancer trial; and in 2030, completion of the adjuvant/resected pancreatic ductal adenocarcinoma trial.

Read more

1. A quintuple agonist in obesity

The obesity space is getting crazier and crazier. This week Eli Lilly reported earnings, with $13B of revenue for Tirzepatide in the first quarter of the year, making it (by far) the highest-selling drug in the world.

To me it seems like GLP1s will become a backbone therapy, a metabolic foundation to be combined with other medicines — or built into multi-agonist molecules — for more weight loss, to reduce side effects, and for benefits beyond obesity.

This paper from Nature this week takes it to the next level, with a molecule that agonises 5 different receptors: GLP1, GIP, PPARα, PPARγ and PPARδ.

GLP1 agonism reduces hunger, while adding GIP tolerises and increases weight loss (this is the mechanism of Tirzepatide). PPARα/γ/δ agonism is how Lanifibranor works, which is itself in phase 3 trials for fatty liver disease — this adds anti-inflammatory effects, lowers liver fat, and increases insulin sensitivity.

Put that all together, and the result is 40% weight loss — when tested in mice!

It is a wild statistic that the two main GLP1s (Semaglutide and Tirzepatide) make more revenue than the two main AI labs, OpenAI and Anthropic, combined. Still, we’re just at the beginning of this wave: the leader in the space, Eli Lilly, has no less than 8 weight loss drugs in its pipeline.

2. A new hair loss drug wins in phase 3

The approved drugs for pattern hair loss are oral finasteride and topical minoxidil. These are decent therapies, but oral finasteride is generally only used by men (it can harm development of the male fetus in pregnant women) and there are often concerns over side-effects (sexual, mood). Minoxidil is usually used topically (due to risk of systemic side effects with oral use) so it’s applied directly to scalp.

This week, VeraDermics, a dermatologist-founded company in Connecticut, posted phase 3 results for an oral, reformulated version of minoxidil. Here’s an example of what it can do:

The reformulation means that release of minoxidil is steadier, and longer-lasting — to keep its levels high enough for continuous hair growth (but not high enough to be toxic).

Read more

If you’re running systematic reviews in pharma, biotech, or academia, check out the AI-native systematic review platform I helped build — scholara.ai!

1. The first FDA-approved gene therapy for genetic hearing loss

About 20-50 newborns per year (in the US) have deafness caused by mutations in the OTOF gene, which causes severe-to-profound hearing loss. OTOF deficiency means that hair cells in the ear can detect sound but can’t transmit the signal to the auditory nerve.

A gene therapy for this condition just got accelerated FDA-approval this week, based partly on this trial. At the start of the trial, hearing loss was ~equivalent to an inability to hear a lawn mower. The patients all had surgery to deliver the functional gene (OTOF is huge, so it’s delivered in two separate vectors, and recombines in target cells inside the ear.)

9 patients (of 12) gained enough hearing to avoid cochlear implants, 6 developed ability to hear soft speech, and 3 reached normal hearing. The best thing is to watch the video, it is quite amazing:

This is life-changing for patients: they started to interact with friends, respond to parents, and say early words.

Here are the aggregate data, which show an objective measure of hearing based on auditory nerve/brainstem activity upon playing a “click”.

Dr. Veera Rajagopal, who worked at Regeneron, put the science that led to this breakthrough together on X — there was a 30 year journey from discovery of the causative gene (OTOF) to this week’s approval.

2. A new medicine for alopecia?

Alopecia areata is a rare cause of hair loss — it’s an autoimmune condition where the immune system attacks hair follicles, with a prevalence of about 1-2 per 1,000 people.
All of the approved drugs are JAK inhibitors, which work great — but they have black box warnings (infections, lymphoma, cardiovascular disease).

Nektar therapeutics have an IL2 receptor agonist, with possibly the coolest name in biotech at the moment: ‘Rezpegaldesleukin’. Unlike most anti-inflammatories, which work by blocking proteins, Rezpeg works by expanding anti-inflammatory T regulatory cells:

In December, Nektar released the first set of data at 36 weeks of Rezpeg treatment, which already looked pretty good (but ultimately missed the primary endpoint).

This week they released longer-term follow-up data, up to 52 weeks, and it looks like the responses actually get better with time.

These are phase 2 data, so it’s still pretty early — the phase 3 trial (which is needed for approval) will start this year. If it succeeds, Rezpeg could be a safer option for alopecia areata that roughly matches the efficacy of (low dose) JAK inhibitors — with first-line potential.

3. The most effective oral cholesterol medicine yet

PCSK9 inhibition is probably the most effective mechanism to lower LDL cholesterol. The idea is built on an amazing natural experiment: people born with inactivating mutations in PCSK9 have low cholesterol from birth, and up to 10 X reduction in their risk of coronary heart disease.

There are already three medicines in this class — Inclisiran, Alirocumab and Evolocumab — all of which are (expensive) injectables. Merck (MSD) have a new, oral PCSK9 inhibitor, Enlicitide, a small peptide that blocks PCSK9. This means PCKS9 can’t degrade the LDL receptor, so there’s more of the receptor, which takes LDL from the blood (where it can cause atherosclerosis) into the liver (where it can’t).

These are some really really nice phase 3 data, published this week in JACC, that pit different cholesterol medicines against each other. In people already taking a statin, Enlicitide was tested against Bempedoic Acid, Ezetimibe, or the Bempedoic Acid-Ezetimibe combination.

Enlicitide lowered LDL more than even the combination, from a baseline LDL of ~90 mg/dL down to the low 30s.

An extra benefit of the PCSK9i class is its effect on lipoprotein(a) — here, it cut the levels by ~25%.

It’s worth noting that there are no cardiovascular outcome data for Enlicitide yet, but its efficacy on LDL cholesterol is ~similar to Evolocumab, which is likely to be reflected in cardiovascular benefit.

Read more

If you’re running systematic reviews in pharma, biotech, or academia, check out the AI-native systematic review platform I helped build — scholara.ai!

1. KRAS is druggable in pancreatic cancer

KRAS is the most commonly-mutated oncogene in cancer — it’s a driver in ~90% of pancreas cancers, ~30% of lung adenocarcinomas, and ~40% of colorectal cancers. KRAS is a small, smooth molecule, which is a problem for drug development, because there’s no obvious pocket or active site for a drug to slot into.

KRAS does sometimes expose a druggable pocket — in its ‘off’ state. There are a couple of approved KRAS inhibitors that do this (Sotorasib and Adagrasib, in colorectal and lung cancer) both of which target the G12C mutant. These are effective, but the G12C mutant only composes a fraction of KRAS mutants, and mutant KRAS spends most of its time in the on state — not the off state.

To drug the ‘on’ state, you’d need a drug small enough, with high enough affinity, to displace GTP. Given the picomolar affinity with which GTP binds to KRAS, and its high concentration in the cell, this is probably ~impossible, so Revolution Medicines came up with a different way.

Their lead compound is Daraxonrasib, a molecular glue which binds to a scaffolding protein (called Cyclophilin A) and then KRAS. This trimer locks KRAS into an inactive state that’s unable to activate the downstream pathway, and promotes GTP hydrolysis — to push KRAS into the ‘off’ state.

This mechanism means that Daraxonrasib inhibits many of the major KRAS mutations — G12D, G12V, G12C, G13X, Q61X, and others:

This week, RevMed press-released their phase 3 results for Daraxonrasib in second-line metastatic pancreatic cancer. The Daraxonrasib group had an overall survival of 13.2 months vs. 6.7 months for the group treated with the physician’s choice of chemotherapy — a hazard ratio of 0.4. To put this in context, no other therapy has ever achieved more than 10 months overall survival in this setting:

The major caveat here is that this is just a press release — the full dataset will be announced at ASCO, in Chicago in late May (I will be there!).

It’s worth mentioning here that WarpSpeed released a prediction for this clinical trial, using AI. They predicted 6.77 months survival in the control arm (spot on) and a pessimistic 10.14 months with Daraxonrasib:

If these data were not exciting enough, this week, at the AACR conference, RevMed released the phase 1/2 data for Daraxonrasib in the first line of metastatic pancreas cancer. They trialled Daraxonrasib alone and in combination with standard-of-care chemotherapy. The data look good there, too — here’s how they compare to the current standard of care:

Beyond the phase 3 2L mPDAC data announced this week, RevMed are running two other phase 3 trials: in 1L metastatic PDAC and in adjuvant resectable PDAC.

Credit to the Gut Onc Lab pod for helping with my research for this!

2. KRAS G12D is druggable in lung cancer

Read more

This newsletter is sponsored by Scholara — the AI-native systematic review platform (ok, it’s my own app, check it out!). If you’d like access for your team, you can try out the full version for free, just send us an email.

1. A more delicate (potentially curative) therapy for sickle cell disease

About 500,000 people are born each year with sickle cell anemia. It’s a genetic condition that causes red blood cells to become misshapen — sickled. It’s also a fascinating evolutionary trade-off: one mutant copy helps protect against malaria; two cause sickle cell disease. These cells block vessels, potentially leading to multi-organ damage if untreated.

The first-ever CRISPR therapy (Casgevy) was approved for sickle cell disease in 2023, which takes patients’ hematopoetic stem cells and edits them to raise levels of fetal hemoglobin and reduce sickling. Casgevy works by disrupting expression of a gene called BCL11A — which encodes a transcription factor that suppresses fetal hemoglobin by repressing two genes, HBG1 and HBG2.

Casgevy isn’t perfect, though: the CRISPR approach introduces double-strand breaks, which can be the substrates of genome rearrangement (this sounds about as bad as it is).

This week, Beam Therapeutics report in the New England Journal of Medicine a more delicate approach: base editing of individual sites in the promoters of HBG1 and HBG2 — to prevent binding of BCL11A. It’s a more native strategy that avoids double-strand breaks, and — so far — it works beautifully:

Some patients even overshot into high hemoglobin (which was not obviously harmful) and the result was an operational success, too: there were no manufacturing failures to derail treatment — often a scourge of hematopoietic stem cell therapies.

This still is not a simple cure, though. One patient sadly died, which was attributed to idiopathic pneumonia, probably caused by the busulfan conditioning that’s a requirement of the hematopoietic stem cell transplant.

That problem is important in its own right — and Beam are already thinking about it in waves: Wave 1 (this one): ex vivo base editing of a patient’s own hematopoietic stem cells to raise fetal hemoglobin. Wave 2: replacing busulfan with gentler, targeted conditioning to make the transplant process safer. Wave 3: in vivo editing of hematopoietic stem cells, aiming to avoid transplantation and conditioning entirely.

Read more

This newsletter is sponsored by Scholara (ok, it’s my own platform) — check it out! If you’d like access for your team, you can try out the full version for free — just send us an email.

1. Multiple myeloma progress

In 1986, the best therapy for multiple myeloma (relapsed or refractory, after 1-3 lines of previous therapy) was a steroid, dexamethasone — and progression-free survival was about 3 months. Since then, sequential breakthroughs mean that now >80% of people can go years without progression, and (the trials haven’t run for long enough yet) but there might be permanent disease control.

These data compare different trials from different times (and between patient groups that are not absolutely equivalent), so they are not directly comparable — but it works nicely as an illustration.

2. Predicting clinical trial success with AI

Can clinical trial success be predicted using AI? These guys have done it, and provided 10 of their predictions. One of them is for Daraxonrasib — the pan-Ras inhibitor that’s currently being tested in RASolute 302, a phase 3 trial vs. chemotherapy, in metastatic pancreatic cancer:

Here’s the prediction: progression-free survival of 5.39 months for Daraxonrasib (vs. chemotherapy’s 3.39 months), which translates to an overall survival of 10.14 months (vs. 6.77 months with chemotherapy).

To do this, they modelled a few things — control arm survival, phase 2 to 3 efficacy translation, progression-free survival to overall survival translation, statistical design and readout maturity — and ran this through 100,000 Monte Carlo simulations.

Here’s some more details on their thesis:

This idea is fascinating to me — I will watch and update how close their predictions are!

3. GLP1s are bigger than AI

Summing the revenue made by GLP1 weight loss drugs (Semaglutide or Tirzepatide) and comparing them to all of the revenue made by the frontier AI labs (OpenAI or Anthropic) GLP1s are way ahead.

As someone on Twitter said, this is a ‘funny if a bit senseless comparison’ but… here it is:

Read more

This newsletter is sponsored by Scholara (ok, it’s my own app) — check it out!

1. The first US-approved weekly insulin

Basal insulin, which provides a baseline level of insulin that can be topped up at mealtimes, is usually given as a once-daily injection. Novo Nordisk’s Insulin Icodec was just FDA-approved this week — the first weekly insulin to be approved in the US — which reduces the burden of weekly basal insulin injections from 7 to 1.

The approval is for type 2 diabetes. In the trial, it was better than daily insulin (Glargine) to control HbA1C. The main potential safety concern of a weekly insulin is dangerously low blood sugar (hypoglycaemia), but there was no clear, statistically significant increased risk of this in the trial.

What about in type 1 diabetes? There are trials there, too, but weekly insulin did lead to increased hypoglycemia compared to daily insulin. This held back the US from approving Icodec for type 1 diabetes, but Canada and the EU both approved it — for both type 1 and 2 diabetes — back in 2024.

It’ll launch in the US in the second half of this year, with expected annual sales of $5B/year by 2030. Looking into the future, Novo Nordisk are also combining Icodec with their lead GLP1 receptor agonist, Semaglutide, together in the same pen — it’s called IcoSema, and it’s already approved in the EU!

2. The lower the LDL cholesterol, the better

It’s clear from genetic evidence, from meta-analyses, and from trials with increasing intensity of cholesterol-lowering therapies, that lower cholesterol tends to better protect against heart disease. It’s been shown previously, too, that an LDL target of <70 mg/dL beats a target of <90 mg-110 mg/dL. But what about an even lower target: <55 mg/dL?

That’s what was tested here, in patients with cardiovascular disease — presented in the NEJM this week, and at the ACC26 conference. The effect is huge: 33% fewer cardiovascular events (defined as death from cardiovascular causes, nonfatal myocardial infarction, nonfatal stroke, any revascularization, or hospitalization for unstable angina at 3 years).

The absolute risk reduction was 3.1% over these 3 years, for a number-needed to treat of 32 (roughly 32 patients need intensive targeting for 3 years to prevent 1 cardiovascular event). The intensive targeting group did this through more use of high-intensity statins, more ezetimibe use, and slightly more PCSK9 inhibitor use.

The striking thing is that the intensive-lowering group only reduced LDL cholesterol by ~15% (56 mg/dL vs. 66 mg/dL). The 33% reduction in cardiovascular events that this led to is more than expected, and it’s unclear why that is — there may be population-specific effects, since the trial was specific to South Korea.

With intensive lipid-lowering, safety concerns would be new-onset diabetes, muscle symptoms, and liver enzyme changes, but none of these were different between the groups. The only safety difference was fewer events of creatinine elevation (i.e. better kidney function) with intensive lipid-lowering.

3. A chemotherapy-boosting therapy gets approval

When I was doing pre-clinical cancer research, we spent a lot of time trying to find new agents to boost the efficacy of chemotherapy. There’s still only one drug approved for this purpose, though: Leucovorin, which boosts the efficacy of fluorouracil chemotherapy.

Corcept therapeutics now have a second, which was just approved this week. It’s a glucocorticoid receptor blocker, which inhibits the cortisol signalling that cancer cells use to resist chemotherapy.

In the trial, in patients with ovarian cancer (resistant to previous lines of chemotherapy), combining Relacorilant with the standard of care chemotherapy, nab-paclitaxel, increased overall survival:

This approach could, in theory, generalise to other chemotherapy backbones, other cancer types, or even immunotherapies too (and Corcept are pursuing each of these).

4. An oral IL23 receptor blocker approved for psoriasis

Read more

Before I start, I’ve been working with some others to build the world’s first AI-native systematic review platform. It helps automate the entire end-to-end workflow (with the researcher in the loop) in hours, not months. If you run systematic reviews, I’d love for you to try it — it’s available for free at scholara.ai. I’ve included some original research we did with it later in the newsletter!

1. How immunotherapy transformed survival in melanoma

This week I realised that GPT-5.4 Pro can help reconstruct and merge Kaplan-Meier curves from different trials, which opens the opportunity to directly show how much progress has been made so far (and how much there is left to do).

Advanced melanoma used to be treated with chemotherapy, with an average survival of <12 months. Immunotherapy means that now more than 50% of patients live for at least 5 years (many with very durable remission).

The curves come from the CheckMate 066 and CheckMate 067 trials, both for first-line treatment. For clarity, I left out some regimens (Pembrolizumab, combinations with anti-LAG3, combinations with BRAF and MEK inhibitors).

2. How EGFR inhibitors changed outcomes in EGFR-mutant non-small-cell lung cancer

Advanced EGFR-mutant non-small-cell lung cancer used to be treated with chemotherapy alone. Gefitinib was the first EGFR inhibitor with a positive phase 3 trial, in 2010, and since then there have been further, stepwise improvements. Osimertinib is today’s most effective EGFR inhibitor, especially when combined with a certain chemotherapy.

To do this, I merged the Kaplan-Meier curves from four trials (1, 2, 3, 4) for first-line treatment of EGFR-mutated (exon 19 deletion or L858R mutation) advanced non–small-cell lung cancer. I left out some intermediate regimens for clarity, but this tells the story of how median progression-free survival was transformed from 5.4 to 25.5 months.

3. Do GLP1s cause pancreatitis?

Semaglutide and Tirzepatide are now the top two-selling drugs globally, but there is still a safety concern over acute pancreatitis. The last systematic review of the randomized evidence is now almost two years old (its search was done in May 2024) and there have since been a lot of new data on this, so I updated it (using scholara.ai).

Across 31 randomized trials with 40,274 patients, here are the results.

In the placebo group, there were 50 cases of acute pancreatitis across 17,433 patients; in the Semaglutide/Tirzepatide groups, there were 59 cases/22,841 patients — the odds ratio was 0.99 (95% CI 0.67-1.45).

The studies included here are: phase 3 randomized trials of drug vs. placebo; subcutaneous semaglutide or tirzepatide only (because these are the modern, approved GLP1s that >95% of GLP1-receiving patients are taking today); only the studies that report acute pancreatitis as an adverse event.

There’s no obvious pancreatitis safety signal in these pooled randomized data (which is reassuring) but the overall incidence of acute pancreatitis was very very low. We also published this as a preprint, if you’re interested.

4. How modern therapies displaced classical chemotherapy in advanced urothelial carcinoma

Urothelial carcinoma is an amazing example of a cancer where classical chemotherapy has been displaced by modern therapies: antibody-drug conjugates and immunotherapy.

To do this I merged the Kaplan-Meiers from three phase 3 trials in first-line locally advanced or metastatic urothelial carcinoma: 1, 2, 3.

Read more

Fasting is somewhere atop hype mountain currently, in my opinion. In its many forms, it has gripped the wellness world and does not seem to be letting go.

Fasting likely doesn’t improve breast cancer treatment

As a researcher in oncology, one specific fasting study caught my attention while investigating fasting. Between 2014 and 2018, Dutch oncologists randomised breast cancer patients to chemotherapy followed by surgery with or without a Fasting Mimicking Diet (FMD). FMD doesn’t strictly compose of a “fast”, but rather a period of reduced caloric intake (often a few hundred calories) over a few days followed by a longer period of normal eating.

Anyhow, in this Dutch trial, it appears that patients adhering to FMD had better responses to their chemotherapy, in their surgery specimens, more FMD-patients had high levels of necrosis in their tumours. Indeed, Miller-Payne grade 4/5, corresponding with over 90% of cancer cells being dead was more common in those adhering to FMD - the per protocol analysis.

Additionally, quite some patients in the control group also fasted! This likely masks some of the potential benefits. These results are similar to those of a small Iranian study published in a Frontiers journal.

Sounds great, so all cancer patients should fast? Maybe not. Compliance to FMD is quite poor, hence the significant differences only occurring in the per protocol analysis, not the intention to treat analysis. More importantly, a planned long-term follow up of these patients was aborted due to futility. Doesn’t sound promising.

Fasting does not lead to more weight loss

In the now almost unimaginable world of pre-GLP1 weight loss, fasting was recommended for weight loss by lay people, professionals and everyone in between. It was easy to miss in the tsunami of GLP- and other peptide related weight loss news, but a Cochrane systematic review and meta-analysis published last month (February 2026) showed that fasting is not better than regular diet advice for weight loss. There wasn’t an effect on HDL, total cholesterol and triglycerides either.

It’s unsure whether fasting helps control auto-immune diseases

But fasting does strengthen the immune system right? You’d hope it does one thing right, but sadly this may also not be the case. Studies found discordant effects of fasting on immune function, likely relating to the strictness of the fast: extreme fasting limits immune function in mice, whereas moderate fasting apparently doesn’t.

In some individuals, fasting may be beneficial, consider for example people with multiple sclerosis (MS), an auto-immune disease affecting white matter in the central nervous system. Here, mice affected with Experimental Autoimmune Encephalomyelitis (EAE), a model for MS, undergoing intermittent fasting have less symptoms than control mice. Keep in mind though, EAE in mice is extremely easy to treat.

In humans, some clues hint that fasting may help MS too. Some (but not all) of the immune cell types implicated in MS pathogenesis drop when patients fast (shown in a small randomised controlled trial). In mice with EAE, fasting increased the immunoregulatory metabolite β-hydroxybutyrate (BHOB), and limits breakdown of myelin sheets as shown below. There is not sufficient clinical data though.

Read more

Before I start: we built a new systematic review platform that helps you search, screen, synthesize studies, generate hypotheses, analyze statistics, run meta-analyses, and draft the paper with colleagues — even for long lists of studies or messy data. It’s fast, too: <4 hours on average for the end-to-end process.

We’re improving it every day. Try it for free, or book a demo, here: scholara.ai.

1. A muscle-sparing weight loss medicine

Despite their meteoric rise, there are still concerns over obesity drugs (GLP1s) — one is that they reduce muscle mass.

This study, a phase 2 trial this week in Nature Medicine, combines the GLP1, Semaglutide, with Bimagrumab, a muscle-boosting antibody initially tested in muscle-wasting conditions.

The combination of Semaglutide + Bimagrumab led to more weight loss, more fat loss, bigger reductions in inflammation and less lean mass loss:

This all comes at a cost, of course: side effects enriched in the combo group were muscle cramps, acne, and nausea/vomiting/diarrhoea.

I think this is particularly interesting because it gives a glimpse into the future of GLP1s: as a backbone therapy. For example, add bimagrumab to preserve muscle; add other peptides (GIP/GCG/amylin) for more weight loss, add a PCSK9 mAb for even lower cholesterol, add to anti-inflammatories to increase their effect (e.g. in Crohn’s disease, psoriatic arthritis), and add to liver-directed therapies (e.g. FGF21 agonists) to increase their effect. Every one these is being pursued, including by Eli Lilly, Innovent Biologics, and Novo Nordisk.

2. Approval of a multiple myeloma regimen with unprecedented efficacy

Multiple myeloma is a malignancy of the antibody-producing plasma cells. The acronym for its signs and symptoms, that you learn in med school, is ‘CRAB’. It causes high Calcium and Bone lesions, because the malignant cells infiltrate bone, and Renal dysfunction and Anemia, because the malignant cells produce abnormal antibodies that damage the kidneys and suppress normal haematopoiesis.

This trial, in the NEJM, tests dual therapy with daratumumab (a CD38 antibody that depletes malignant plasma cells) and teclistamab (a bispecific antibody that binds CD3 on T cells and BCMA on multiple myeloma cells, bringing them together and helping the T cells kill the cancer cells.). This regimen already has FDA-approval in multiple myeloma, but only after 4+ lines of previous therapy — this trial moves it forward, testing it as second-line therapy.

These data were presented initially on the 9 Dec 2025, as a late-breaker at American Society of Haematology. Shortly after, on 15 Dec 2025, the regimen received an FDA Priority Voucher, and was FDA-approved this week, on 5 Mar 2026: 4 months after data!

The Kaplan Meier curve shows the incredible progress that it represents:

This doesn’t come without its own risks, though: cytokine release syndrome and hypogammaglobulinemia are both adverse events, so close monitoring and infection prevention measures are required.

3. Remission of autoimmunity with an experimental antibody

Recently there have been some amazing reports that Blinatumomab, a B cell-depleting antibody, can induce remission of autoimmune disease — rheumatoid arthritis, systemic sclerosis, myaesthenia gravis — in patients whose disease responds to no other therapy.

Blinatumomab works in a similar way to Teclistamab (discussed above) — it’s a bispecific CD3xCD19 T cell engager, which deeply depletes the B cell pool by linking B cells (which produce aberrant antibodies) to T cells (which kill them).

Another case was published this week in NEJM, which described a 31-year old woman with a disorder of both bleeding and thrombosis (combined immune thrombocytopenia and antiphospholipid syndrome), which was resistant to every therapy tested — glucocorticoids, rituximab, fostamatinib, azathioprine, daratumumab, even splenectomy.

After all of this, Blinatumomab was tried, and was able to induce disease remission: the B cell pool was reset, autoantibodies were depleted, platelet levels recovered, and bleeding resolved, with no further thrombosis.

There’ve still been no big trials that test Blinatumomab in autoimmunity yet, but its potential to reset the immune system is extremely promising.

4. More evidence that GLP1s help with addiction

If you talk to people taking GLP1s, they’ll tell you they’ve not only stopped eating less, but they might’ve started drinking less alcohol — and maybe they’ve even stopped smoking, too.

This might be because GLP1s seem to calm the brain’s drive to binge:

Read more

Before I start: if you run systematic reviews in academia, biotech, or pharma, we built a new platform that’ll help you do the end-to-end workflow (question to meta-analysis) — in minutes to hours.

We’re improving it every day. Try it for free: scholara.ai.

1. A small molecule to re-activate p53 in cancer

p53 is the ‘guardian of the genome’ — it’s a tumour suppressor that’s mutated in about half of cancers (and deregulated in most of the others). p53 is the reason elephants get much less cancer than they’d be expected to, given their size — they have 20 copies of it!

Re-activating p53 has always been a moonshot goal in cancer therapy, but it’s almost always failed too. To sum it up, here’s a a brief history:

2003 — Gendicine, a p53 gene therapy, approved in China (but not elsewhere)
2012 — The first small molecule MDM2 inhibitor (RG7112, which works by preventing degradation of p53) enters trials, but did not progress to approval
2013 — Second-generation MDM2 inhibitor (Idasanutlin) reached phase 3, but failed and was dropped
2017 — A dual MDM2/MDMX inhibitor, ALRN-6924, enters trials; it has not since been approved
2020–2021 — A mutant p53 reactivator, Eprenetapopt, fails in phase 3
2023 — An MDM2 inhibitor, Milademetan, fails in phase 3
2024 — Promising phase 1 data for an MDM2 degrader

This week, PMV Pharma released phase 1 data in the New England Journal of Medicine. It’s an oral p53 reactivator, called Rezatapopt, which fits into a pocket exposed by p53’s Y220 mutation, re-folding it to look more like wildtype p53. This mutation is surprisingly common — it affects ~1% of all solid tumours.

Across a range of tumour types, all refractory to many rounds of previous therapy, the overall response rate (meaning complete or partial response) was 20%:

One concern with p53-reactivating therapies is the extremely strong selection pressure cancers exert to inactivate p53 (and there are thousands of ways to do it) — for example, straightforward resistance mechanisms here could be MDM2 overexpression (to speed up p53 degradation) or just mutation of a different residue. To what extent this happens with Rezatapopt, and how fast, isn’t yet known.

With recent breakthroughs for inhibitors of RAS (the most commonly mutated cancer gene) and now these early data for a p53 re-activator (the second most mutated) — this is a very very exciting time in Oncology.

2. The most effective weight loss drug yet?

The first-generation obesity drugs (like Semaglutide) were single agonists of the GLP1 receptor. The second-generation drugs (Tirzepatide) were dual agonists, for both GLP1 and GIP receptors. The third-generation is now here, and we’re calling them ‘Triple Gs’ — they’re triple agonists of the GLP1, GIP, and glucagon receptors.

The first triple G, Eli Lilly’s Retatrutide, is ridiculously effective:

It turns out that Novo Nordisk, Eli Lilly’s main competitor, have been working on their own triple G — it’s called UBT251, and they released the phase 2 data this week. Here’s how it compares to Eli Lilly’s Retatrutide at the same 24 week timepoint — even faster weight loss!

There are caveats to this comparison: it’s only phase 2 data, in different populations (UBT251 was tested exclusively in a Chinese population) — and there’s no precise safety data for UBT251.

If it holds up in the phase 3, UBT251 could be particularly useful for those with very high BMI (>40) or those unable to lost weight with other drugs (and it’ll need a different name, too!).

3. Stem cells help treat frailty

Something we’ll almost all face as we get older is frailty, and we are going to need some serious innovation to address it.

This study, published this week in Cell Stem Cell, is an extremely interesting randomized trial of stem cells in frail older adults, aged 70-85. The study measured 6-minute walking at the beginning, and then up to 9 months after stem cell infusion. There were 30 patients in placebo group, and 118 patients in stem cell groups.

The data are really nice. As you’d expect, the walking distance declines in the placebo group, and there’s some dose-dependence, too: at the high dose, stem cells boosted walking distance by about 20%.

So how does it work? The stem cells (donor-derived mesenchymal cells) seem to act mainly by releasing signalling factors to reduce inflammaging and improve function of blood vessels (they do not replace worn-out tissues directly).

4. A first-in-human trial in spina bifida

Spina bifida affects about 1 in 2,875 births in the US (about 1,278 babies per year) — it’s a birth defect where the baby’s spine and spinal cord don’t develop properly early in pregnancy, leaving a gap in the spine.

Read more

Try it for free here: scholara.ai.

It was a busy week of breakthroughs! Let’s get into it…

1. Early stages of a breast cancer vaccine

There’s still only one therapeutic cancer vaccine that’s ever won FDA-approval — it’s this one, for prostate cancer. Other than this, the most advanced trial is for Moderna’s melanoma vaccine.

More recently there’s been excitement over personalized vaccines, where tumors are removed by surgery and vaccines are generated against tumors’ specific antigens (the ‘neoantigens’) to train the immune system and prevent the cancer from returning. Examples include early trials for vaccines in kidney and pancreatic cancer.

This is a new trial, in Nature this week, in triple-negative breast cancer. This type of breast cancer lacks any of the three receptors (the estrogen receptor, progesterone receptor, or HER2) that can be targeted with currently effective drugs — so new therapies are badly needed.

The trial’s just small — 14 patients (and no control group!) — but 11 of the vaccine recipients were relapse-free at 6 years:

BioNTech funded this trial — they’re also running trials for off-the-shelf vaccines (in melanoma and non-small cell lung cancer), in HPV-positive head and neck cancer, and for other personalized vaccines in colorectal cancer, bladder cancer, and pancreatic cancer.

2. The first approval for locally-advanced pancreatic cancer in >30 years

Another very welcome breakthrough was in (locally-advanced) pancreatic cancer, where there’ve been no new therapies for >30 years. That changed this week with the approval of Novocure’s tumor treating fields — delivered via a wearable device, like this:

For the technical details: tumor-treating fields deliver low-intensity electric fields that disrupt microtubule assembly to cause mitotic catastrophe (and apoptosis) in dividing cells. The fields also seem to impair DNA repair, so can sensitize cells to DNA-damaging chemotherapy.

The survival advantage is admittedly modest, but significant:

This is the third approval for tumor treating fields, following glioblastoma and mesothelioma.

3. An update to the standard of care in bladder cancer

The standard of care in muscle-invasive bladder cancer is the 70-year old chemotherapeutic, cisplatin (plus surgery). The issue is that about half of patients are ineligible for cisplatin, mainly because of poor kidney function — which can happen as a direct result of the bladder cancer. For these patients, there’s no clear standard-of-care.

This trial tests the antibody-drug conjugate, Enfortumab Vedotin (which directs a toxic compound, monomethyl auristatin E, selectively to bladder cancer cells via the Nectin-4 antibody, Enfortumab) — in combination with the PD1 inhibitor, pembrolizumab — combined with surgery, in patients ineligible for cisplatin.

Here are the results, from this week in NEJM: dramatically better event-free survival (74.7%, vs. 39.4%) and overall survival (79.7%, vs. 63.1%) vs. surgery alone:

Here are some of the takes from oncologists around the world on this trial…

and my favourite:

So it’s impressive, for sure, but there’s still some way to go: many patients still didn’t respond, and (as you might expect with a combination of two cancer drugs) adverse events were common.

4. The first therapy for myotonic dystrophy…?

The pathology of myotonic dystrophy type 1 is interesting: it’s caused by a trinucleotide expansion (like Huntington’s) but unlike Huntington’s it’s an RNAopathy, not a proteinopathy.

Toxic RNA transcripts from inherited, expanded repeats in the DMPK gene accumulate in muscle cells, disrupting alternative splicing of many genes and causing problems across the body: muscle weakness, cataracts, cardiac dysfunction. There are no current therapies.

Read more

The study — a systematic review and meta-analysis — was done in 2022, so I thought it was ready for an update! I ran the update today, from scratch, using the systematic review platform I built with other doctors and AI engineers (scholara.ai).

My new analysis included 22 randomized trials, with a total of 2,705,711 participants across 8 screening tests. Here’s what I found:

This means the only screening test with evidence (from randomized trials) of increased lifespan is sigmoidoscopy, with an average boost of 110 days.

I find this interesting for three reasons:

1. This analysis captures net impact: potential for increased lifespan (by finding cancers earlier) but also potential for life years lost (due to complications of screening or its follow-up).

2. The negative findings are at least partly a fault of the primary trials: it’s known that detecting improvement in overall survival with cancer screening requires huge populations or long follow-up times (though the synthesis/meta-analysis approach taken here should help with that).

3. One reason screening might not extend overall survival is because people die from other causes first — like heart disease. Progress in these other areas could start to reveal the overall survival benefit — especially given the evidence that most of these tests do reduce cancer-specific mortality.

The original analysis took ~weeks-months — mine took about 2 hours. You can try our platform for free here.

¹If you’re interested in the included trials, they are as follows:

Endoscopy (for early detection of esophageal cancer in people with Barrett’s esophagus): 1

PSA (for early detection of prostate cancer): 2, 3, 4, 5, 6

Sigmoidoscopy (for early detection of colorectal cancer): 7, 8, 9, 10, 11

Low dose lung CT (for early detection of lung cancer in smokers): 12, 13, 14

Colonoscopy (for early detection of colorectal cancer): 15

Mammography (for early detection of breast cancer): 16, 17

Fecal occult blood test (for early detection of colorectal cancer): 18, 19, 20, 21

EBV testing (for early detection of nasopharyngeal carcinoma): 22

Bacteria and tumours go way back

The research on bacteria and tumours is rather extensive. Think about how the bacterium H. pylori, which causes stomach cancer and lymphomas in the lymphoid tissue of the stomach. In fact, almost 5% of all cancers worldwide (excluding non-melanoma skin cancers) are caused by this pathogen!

Consider also the late 19th century hype of injecting bacteria in patients with advanced cancer. While at the time it was a bit controversial, William B. Coley defended the practice in the March 1897 volume of The British Medical Journal. According to his writing, he personally treated 200 sarcoma patients with only two patients succumbing to the treatments and achieving tumour disappearance or shrinkage in many.

Considering this background, it is compelling to investigate how bacteria specifically interact with tumours.

Untangling noise and signal

We always consider that beyond “outer layers” of our body such as skin, the gastrointestinal and the urogenital tract, we’re perfectly sterile. Not entirely true, as even brushing teeth causes bacteria to invade the bloodstream. Cancer patients, which often have a suppressed immune system and breaches of those outer layers following surgery or radiotherapy may be even more prone to having bacteria in odd locations.

However, cancer patients aren’t that immunosuppressed, reliable mechanisms, such as the complement system and the blood-filtering liver and spleen remain effective. In the absence of marked infection, this implies that if there are any bacteria at all in normally sterile places of our body, they should be scant.

Incidentally, scientists have specialised in detecting exceedingly rare events. A PCR test, amplifying and detecting genetic material to infer the presence of a pathogen can detect as low as 3 copies of genetic material per sample. However, quantifying how often a certain DNA- or RNA sequence is present in a sample gets harder the rarer it is: noise quickly drowns out the signal. This was the exact problem in an important Nature study.

A consequential study and its retraction

In 2020, Nature published an article which assessed the presence of microorganisms in tumours using transcriptomic data from The Cancer Genome Atlas. Results were astounding: 7.2% of all bits of genetic material (called “reads”) in human tumours were of non-human origin! A third of these appeared to belong to fungi, archaea or viruses. The authors then constructed a model using the genetic profile of DNA from microorganisms in human blood samples to diagnose cancers, suggesting an accuracy of almost 100%.

Sadly, these results are as impressive as they are wrong. Red flags were quickly noted by other researchers: microorganisms identified in the Nature paper, supposedly residing in the analysed tumours, included bacteria infecting crustaceans, and plant viruses, this is the result of large-scale misclassifications of which “read” belongs to which organism. Further analyses emphasise the danger of using transcriptomic analyses on low-biomass samples, such as a supposed tumoral microbiome: over-analysed noise appears as a meaningful signal.

The study was retracted four years after publication. However, in the meantime some of the researchers involved started up a biotech company, Micronoma, aimed at developing a plasma microbiome test to identify cancer. They even raised $6.5 million in funding. However, investor enthousiasm couldn’t turn around flawed science, and the company folded some years later.

The revival of intratumoral bacteria

Despite setbacks, the story of bacteria in tumours is not over. In January 2026, Nature Cancer published a study assessing the presence of bacteria specifically in head and neck cancers and correlating this with immunotherapy outcomes. Initially, the authors used bulk RNA sequencing complementing it with 16S rRNA sequencing. This last technique only reads bacterial genetic material and allows for identification and quantification of bacteria.

By stringently analysing the bulk RNA sequencing data and using high-quality references to discriminate human and microbial signals they assessed the amount and types of bacteria in tumour samples. However, this leaves the possibility that environmental bacteria were sequenced too, even if they weren’t originally residing in the tumour.

To filter out such contamination, they used sham samples and performed 16s rRNA sequencing on these. Indeed, these samples, that were not supposed to be colonised by bacteria, contained RNA signatures of Burkholderia, Pseudomonas, and Acinetobacter. By subtracting the signal in these sham samples from the bulk RNA sequencing data, they increased the likelihood that microbial signals in their data effectively originate from inside the tumour.

Further splitting patients in a tumour bacterial burden (TBB) high versus low group using the methodology above, stark differences in immunotherapy therapy appeared. The figure below shows the difference in progression-free survival depending on TBB status.

Read more