#glioblastoma
$NWBO #DCVax #DCVaxForBraelyn 𝗦𝗮𝗺𝗲 𝘃𝗮𝗰𝗰𝗶𝗻𝗲. 𝗦𝗮𝗺𝗲 𝗯𝗼𝗼𝘀𝘁𝗲𝗿. 𝗦𝗮𝗺𝗲 𝗞𝗲𝘆𝘁𝗿𝘂𝗱𝗮. 𝗢𝗻𝗹𝘆 𝘁𝗵𝗲 𝗼𝗿𝗱𝗲𝗿 𝗰𝗵𝗮𝗻𝗴𝗲𝗱. Vaccine first: 555 days. #Keytruda first: 271 days. Once #glioblastoma comes back, patients live six to ten months. The Keytruda-first arm, at 8.9 months, landed right where this disease always lands. Vaccine first: 18.2 months. Nothing close has been reported in a randomized trial of recurrent glioblastoma. A 2021 Cochrane review of the trials at first recurrence found medians between 5.5 and 12.6 months. The few higher numbers anywhere come from single-arm or retrospective studies that choose their own patients. This one was randomized. The reason fits in three lines. Keytruda takes the brakes off the immune system's killers. The vaccine teaches them what to kill. Teach first, then release. 𝗧𝗵𝗲 𝗱𝗲𝗰𝗶𝘀𝗶𝗼𝗻 𝘁𝗵𝗮𝘁 𝗺𝗮𝗱𝗲 𝗶𝘁 𝗽𝗼𝘀𝘀𝗶𝗯𝗹𝗲 𝘄𝗮𝘀 𝘁𝗮𝗸𝗲𝗻 𝗶𝗻 𝟮𝟬𝟬𝟮, 𝗯𝘆 𝘁𝘄𝗼 𝗺𝗲𝗻 𝗮𝗹𝗺𝗼𝘀𝘁 𝗻𝗼𝗯𝗼𝗱𝘆 𝘄𝗮𝘀 𝗹𝗶𝘀𝘁𝗲𝗻𝗶𝗻𝗴 𝘁𝗼. 𝗧𝗵𝗲 𝘁𝗿𝗶𝗮𝗹 UCLA reported it for EANO 2026 (NCT04201873): 16 patients randomized to the vaccine first, 19 to Keytruda first, all with the more aggressive, IDH wild-type form. Hazard ratio 0.32. The two groups did not differ significantly in how much tumor surgery left behind, or in mutation load. The booster in both arms is poly-ICLC, sold as Hiltonol. 𝗛𝗼𝘄 𝗶𝘁 𝘄𝗼𝗿𝗸𝘀 Keytruda kills nothing. It takes the brakes off T cells, the immune system's killers. That only helps if the killers already know what to attack. Teaching them is the job of the dendritic cell. It works like a switch. On, in what immunologists call the type-1 state, it gives the order to kill. Off, it goes silent, and a tumor is very good at holding it off. The vaccine is made of the patient's own dendritic cells, grown outside the body where nothing has switched them off, then loaded with the patient's own tumor. It arrives with the switch already on. And it reads the whole tumor, not a target someone chose. In 2008 one dose of this kind of vaccine, loaded with a man's own tumor, found a virus hiding in it that nobody knew was there. The cells found it because they were not looking for anything. Vaccine first, and the cells teach, and then Keytruda releases an army that knows its target. The trial saw it: more T cells inside the tumor, carrying interferon and granzyme B, the enzyme a killer T cell uses to destroy its target. Keytruda first: chronic inflammation instead. 𝗣𝗿𝗼𝗼𝗳 𝗯𝘆 𝗺𝗲𝗰𝗵𝗮𝗻𝗶𝘀𝗺 The FDA's plausible mechanism pathway, set out by Vinay Prasad and Commissioner Marty Makary in the New England Journal of Medicine, asks five questions. This treatment answers all five. 1. A specific broken link? The dendritic cell's switch, held off by the tumor. 2. A product aimed at it? The patient's own dendritic cells, switched on outside the body. 3. A well-known untreated course? Six to ten months, measured in hundreds of trials. 4. Proof the target was hit? T cells in the tumor carrying interferon and granzyme B. 5. A better clinical course? 555 days against 271. The FDA's guidance adds one more test: the improvement must not come from other treatments or from the disease's natural variability. This trial rules out both. Same treatments in both arms. Randomized, balanced, p below 0.0006. The only thing that changed was whether the instruction came first. That is proof by mechanism. 𝗪𝗵𝗼 𝗯𝘂𝗶𝗹𝘁 𝗶𝘁 The UCLA vaccine is ATL-DC, the academic version of DCVax-L. DCVax-L is Northwest Biotherapeutics', and it was invented by Alton Boynton and Marnix Bosch. Boynton filed on dendritic cells loaded with a patient's own tumor in 1995, three years before he co-founded the company. Bosch joined in 2000 and built the manufacturing. Linda Liau, now chair of neurosurgery at UCLA, led the phase 3. Boynton and Bosch are the named inventors on the patents Northwest holds, and the last two authors on the papers she leads. Between 2002 and 2005 they made three decisions, each against where the field was going: 1. 2002: use the whole tumor, not a chosen target. 2. 2003: add interferon gamma, so the cells come out switched on. 3. 2005: stop the cells halfway and inject them straight into tumors that cannot be removed. The second is what makes the other two safe. The third became DCVax-Direct, for tumors a surgeon cannot remove. Both products carry the whole tumor. DCVax-L loads it in the lab, from the tumor the surgeon took out. DCVax-Direct loads it inside the patient. Wherever there is a tumor, they load the tumor. 𝗪𝗵𝘆 𝗶𝘁 𝗶𝘀 𝘀𝗮𝗳𝗲 In DCVax-L's phase 3, patients were given their own brain tumors back, ground up and loaded onto their own cells. Seven of 331 patients had a serious side effect even possibly related to the vaccine. No autoimmune reactions. 𝗧𝗵𝗲 𝗱𝗿𝘂𝗴𝘀 𝘁𝗵𝗮𝘁 𝗻𝗲𝘃𝗲𝗿 𝗳𝗮𝗶𝗹𝗲𝗱 In 2020 a drug from the same class as Keytruda was tested on its own in recurrent glioblastoma and lost. Six years later the same class sat inside the arm that doubled survival. The drug had not changed. This time the switch was on when it arrived. We may already own many of the treatments we need. We have been giving them with the switch off. 𝗦𝘁𝗶𝗹𝗹 𝘄𝗮𝗶𝘁𝗶𝗻𝗴 DCVax-L's phase 3: 13.0 percent alive at five years against 5.7. DCVax-L has not been turned down by the FDA. It has not been submitted. In the UK, a review meant to take 150 days passed 1,000 in September 2026. Boynton is still chief scientific officer. Bosch is still chief technical officer. 𝗕𝗿𝗮𝗲𝗹𝘆𝗻 Braelyn Hollensbe was sixteen when they found it. She had surgery, and it came back. Surgeons at more than one major center looked at what was left and said no. Linda Liau said yes, and took about ninety percent. Hers is the kind of glioblastoma standard chemotherapy works least well against. Then her family started flying from South Carolina to London for DCVax-L, the vaccine Boynton and Bosch invented and Northwest builds from a patient's own tumor, because it is not licensed where she lives. Three induction doses, then a fourth, fitted around the start of college. Then the Keytruda. In that order. On August 28 a scan went to Liau and the UCLA tumor board. What had been left behind at surgery was not there. On September 15 her mother wrote the sentence she had been waiting to write since Braelyn was sixteen: no evidence of disease. Everything above is an argument about order. For Braelyn it was a schedule. Flip the switch. Then amplify. The video walks through the whole mechanism.
Andrew Caravello, DO159,566 просмотров • 19 дней назад
Last 3 wks toughest yet. Since my #glioblastoma treated in June 2023, having freq partial seizures (usually cold peripherally & nauseated) (up to 10x/day). Difficult to manage. 3 times loss of consciousness. No seizures since latest ⬆️ antiepiletics but vv tired & lethargic.
Professor Richard Scolyer AO90,394 просмотров • 2 лет назад
Next #brain #MRI scan: Friday 27th September 2024 (>16 months post #GBM coming to light in #Poland in May 2023). MRI looking for: 1. #glioblastoma recurrence 2. #aneurysm stability 3. Treatment effects How am I travelling/feeling: ok but more anxious this week (scanxiety worse than ever)! Fingers crossed🤞scan ok! AusoftheYearAwards Sydney Swans
Professor Richard Scolyer AO72,671 просмотров • 2 лет назад
I’ve had #neoadjuvant #immunotherapy for my #glioblastoma (G4, IDH WT, multiple adverse molec prog features) & tomorrow I’m heading to RPAH/Sydney Local Health District for definitive resection! Feeling good & most grateful for incred support from amazing family, friends, colleagues. Professor Georgina Long AO (Bluesky @gvlongphdmd)
Professor Richard Scolyer AO103,248 просмотров • 3 лет назад
Today is surgery day. It’s an uncertain path, but one I’m travelling with support & guidance from family, friends & colleagues worldwide. #braincancer #glioblastoma Melanoma Institute Australia Sydney Health Sydney Local Health District Charles Perkins Centre
Professor Richard Scolyer AO75,709 просмотров • 3 лет назад
$NWBO #𝗗𝗖𝗩𝗮𝘅-𝗟: 𝗧𝗵𝗲 𝗘𝘃𝗶𝗱𝗲𝗻𝗰𝗲 𝗶𝗻 𝗣𝗹𝗮𝗶𝗻 𝗧𝗲𝗿𝗺𝘀 A short, plain-language reading of the survival evidence for DCVax-L in #glioblastoma, and what it means under the MHRAgovuk guideline on external control arms. 📊 𝗣𝗔𝗥𝗧 𝗢𝗡𝗘: 𝗪𝗛𝗔𝗧 𝗧𝗛𝗘 𝗧𝗥𝗜𝗔𝗟 𝗙𝗢𝗨𝗡𝗗, 𝗔𝗡𝗗 𝗪𝗛𝗬 𝗜𝗧 𝗠𝗔𝗧𝗧𝗘𝗥𝗦 DCVax-L more than doubled five-year survival, and the benefit is durability: a subset gets lasting disease control and simply stays alive. 💉 𝗪𝗵𝗮𝘁 𝘁𝗵𝗲 𝗱𝗿𝘂𝗴 𝗶𝘀 DCVax-L is a personalized cancer vaccine for glioblastoma, the deadliest form of brain cancer and a designated orphan disease. It is a living drug: its active ingredient is the patient's own immune cells, primed with proteins from that patient's surgically removed tumor, so the immune system learns to attack the cancer. Unlike a chemical drug that is metabolized and cleared, it switches on a living immune response that keeps working long after the injection. Glioblastoma comes back in almost everyone: even with the full standard of surgery, radiation, and temozolomide chemotherapy, most patients live under two years, and only about one in twenty reaches five. DCVax-L is given on top of that standard care, not in place of it: every patient in the trial received surgery, radiation, and temozolomide, and the vaccine was added to it, so the comparison measures what the vaccine adds. For two decades, nearly every new drug tried in this disease has failed. That is the backdrop against which any positive result must be judged. 📈 𝗪𝗵𝗮𝘁 𝘁𝗵𝗲 𝘁𝗿𝗶𝗮𝗹 𝗳𝗼𝘂𝗻𝗱 In a Phase 3 trial of 331 patients, those who received DCVax-L lived longer. At the median the gain looks modest, about three months (19.3 versus 16.5). The number that matters sits at the far end of the survival curve: more than twice as many vaccine patients were alive at five years, 13.0% versus 5.7%, and a few reached ten years in a disease that usually kills within three. In patients whose tumor had already returned, the effect was larger still, cutting the risk of death by about 42%. 👥 𝗘𝘃𝗲𝗿𝘆 𝗴𝗿𝗼𝘂𝗽 𝗯𝗲𝗻𝗲𝗳𝗶𝘁𝗲𝗱, 𝗲𝘃𝗲𝗻 𝘁𝗵𝗲 𝗵𝗮𝗿𝗱𝗲𝘀𝘁 𝘁𝗼 𝘁𝗿𝗲𝗮𝘁 The benefit was not confined to the easy cases. Of the six prespecified subgroups the trial examined, every single one favored DCVax-L, and there was no group in which it did worse than standard care. The largest gains came in patients whose tumors carry MGMT methylation, who reached a median survival of 30.2 months from randomization against 21.3 for the controls. But even the hardest-to-treat patients, whose tumors lack that methylation, resist standard chemotherapy, and carry the worst prognosis in this disease, still came out ahead with the vaccine, at a hazard ratio of 0.93. A treatment that helps across the whole population, and helps most where the biology is most favorable, is acting like a real drug. 📉 𝗪𝗵𝘆 𝘁𝗵𝗲 𝗺𝗲𝗱𝗶𝗮𝗻 𝗵𝗶𝗱𝗲𝘀 𝘁𝗵𝗲 𝗿𝗲𝗮𝗹 𝘀𝘁𝗼𝗿𝘆 That five-year number is the whole story, and the median buries it. Almost every treatment that ever helped in glioblastoma did the same modest thing: it slid the survival curve a few months to the right, then let it fall back. Doubling five-year survival is different in kind. A three-month gain at the median cannot, by itself, double the fraction alive at five years. The only shape that produces both numbers is a split: most patients get the small delay the median measures, while a subset gets durable disease control that lasts for years, well past where glioblastoma should have ended them. That subset is the long tail of the curve, and it is where the benefit lives. A median ignores extremes, the way a town's median income tells you nothing about its millionaires. The real story is not a longer delay; it is that a meaningful share of patients simply stay alive. 🧬 𝗪𝗵𝘆 𝘁𝗵𝗲 𝗯𝗶𝗼𝗹𝗼𝗴𝘆 𝗽𝗿𝗲𝗱𝗶𝗰𝘁𝘀 𝘁𝗵𝗶𝘀 𝘀𝗵𝗮𝗽𝗲 The tail is not luck. It is what this biology is built to produce. The vaccine carries proteins from the patient's own tumor, so it aims the immune system at whatever that tumor is made of, not a fixed short list of targets. And it amplifies: one trained immune cell drives many others that multiply into cancer-killers. Andres Salazar, the neurologist who developed poly-ICLC into the clinical adjuvant given with the vaccine, puts it in a line: you start the fire, and you keep it burning. A response like that does not produce a one-time bump. It builds and widens over time. What matters is not just that a response forms, but what kind. Dendritic cells are the immune system's master switch for that, the cells that set what kind of attack the body mounts, and this vaccine is built from them. It drives what immunologists call a type 1 polarized response: an interferon-driven, cytotoxic program aimed squarely at the tumor. That direction comes from the vaccine itself; the poly-ICLC adjuvant given with it drives the same interferon program and sustains it. That is the active ingredient, and it has been measured. In UCLA studies of this approach, the patients whose immune systems mounted the strongest interferon response lived the longest. The response also spreads. In a different cancer, a vaccine carried on this same poly-ICLC adjuvant drove more than 70% of a patient's cancer-killing T cells to target proteins that were never in the vaccine: the immune system outgrew its original targets and went after the rest of the tumor on its own. This is called epitope spreading, and it is not particular to one tumor; it is what this kind of response does, and it is the kind of response DCVax-L builds. That breadth is the likeliest thing separating the long-term survivors from everyone else, a response that breaks past its targets and clears the disease rather than one that stays caged and stops. There is even a tell in who benefits most: the effect is largest in tumors whose biology builds up more mutations under chemotherapy, and more mutations mean more targets for a whole-tumor vaccine to find. The same logic explains what the vaccine is not, and what it does not need. Checkpoint inhibitors, the drugs that release the immune system's brakes, have failed on their own in glioblastoma because there was no active response to release; the vaccine supplies that response first. That makes the vaccine the natural foundation for combination therapy. Combining it with its poly-ICLC adjuvant, made by Oncovir, has already shown meaningful survival gains in a published analysis. In the pivotal trial, the vaccine's proven benefit came added on top of standard chemotherapy and radiation; the newer question is how much the immune response can carry on its own. A UCLA trial is now testing that in patients whose tumors have returned, adding #Keytruda (pembrolizumab), the checkpoint antibody from $MRK, in a regimen built entirely around the immune response with no chemotherapy or radiation in it at all. Its interim survival curve shows the shape the biology predicts. In the arm given the vaccine and Keytruda together after surgery, the curve does not fall away but flattens into a plateau, with roughly 65% of patients still alive well past the point where recurrent glioblastoma kills nearly everyone. That plateau is the signature of a response that took hold and lasted, the type 1 attack forming durable immune memory so that once the disease is controlled it stays controlled. That is where this points, and where it is already arriving, a treatment that works through the response itself, one that could in time lean less on the harsh radiation and chemotherapy that have defined glioblastoma care and barely moved its survival. These are interim results, from Prins, Cloughesy, and Liau at UCLA. 🔍 𝗣𝗔𝗥𝗧 𝗧𝗪𝗢: 𝗜𝗦 𝗜𝗧 𝗥𝗘𝗔𝗟? The trial was randomized, an independent experiment shows the outside comparison is trustworthy, and every separate check points the same way. 🤔 𝗧𝗵𝗲 𝗼𝗯𝗷𝗲𝗰𝘁𝗶𝗼𝗻, 𝗮𝗻𝗱 𝘄𝗵𝘆 𝗶𝘁 𝗺𝗶𝘀𝘀𝗲𝘀 𝘁𝗵𝗲 𝗺𝗮𝗿𝗸 So much for what happened; the harder question is whether to believe it. The trial has the feature critics attacked: it could not keep a normal placebo group, because patients assigned to placebo were allowed, by design and by medical ethics, to switch to the vaccine once their cancer returned, and almost all did. That erased the internal comparison, so survival was measured against closely matched patients from other completed trials, an approach called an external control, which critics argued could tilt toward the vaccine. What the objection misses is where the randomization went. This was a randomized, blinded trial. The patients who got the vaccine were assigned to it at random, not hand-picked, so the treated group is an ordinary slice of the trial population, not a favorable one. The crossover removed the placebo group but never touched how patients were assigned. That leaves exactly one place for bias to enter, the outside comparison group, which is precisely what the next checks test. ✅ 𝗧𝗵𝗲 𝗰𝗵𝗲𝗰𝗸 𝘁𝗵𝗮𝘁 𝗺𝗮𝗸𝗲𝘀 𝗶𝘁 𝘁𝗿𝘂𝘀𝘁𝘄𝗼𝗿𝘁𝗵𝘆 Before trusting a scale to weigh something unknown, you confirm it reads zero with nothing on it. That is what the calibration does, and it is the strongest part of the case. A separate, independent randomized trial called INSIGhT was run through the very same external-control method, and it gave two answers. First, three experimental drugs that had already failed were run through it, and it correctly found nothing (hazard ratios of 1.00, 0.93, and 0.88): the method does not manufacture a benefit where none exists. Second, INSIGhT's external controls were set head to head against its own randomized internal controls, and they were statistically indistinguishable. In this disease, an external control reproduces the answer a real randomized control would have given, and because that trial belonged to a different group, no one can say a sponsor graded its own work. The one place bias could enter, the control side, is the one place an independent randomized experiment certified as clean. There is a deeper fit worth naming, and it is what makes the two halves of this case one. The same instrument that reported those three failures as failures reads the vaccine as a success, and the biology says why: those drugs could not hold a tumor this varied, and the vaccine builds the broad, lasting response that finally does. One method, opposite readings, and one mechanism behind both. The statistics and the biology are not two arguments. They are the same argument seen twice. 🃏 𝗪𝗮𝘀 𝘁𝗵𝗲 𝗰𝗼𝗺𝗽𝗮𝗿𝗶𝘀𝗼𝗻 𝘀𝘁𝗮𝗰𝗸𝗲𝗱 𝗶𝗻 𝘁𝗵𝗲 𝘃𝗮𝗰𝗰𝗶𝗻𝗲'𝘀 𝗳𝗮𝘃𝗼𝗿? A natural worry is that the outside comparison was arranged after the fact to flatter the vaccine. The trial was built to prevent that. The patients to compare against, and the rules for matching them, were fixed in writing before anyone saw results, and an independent firm, not the company, chose the comparison trials against those rules. The trial also switched its main measure partway through, from delaying tumor growth to overall survival, but that was not a maneuver: immune treatments cause a harmless swelling that mimics tumor growth on scans and made the growth measure unreliable, and the switch was made while everyone was still blinded. Three further checks point the same way. Survival in this disease has not improved over the years the comparison spans, so same-era controls are sound. When the borrowed controls were tested directly, the comparison came out conservative rather than flattering. And the controls were counted from the same point in the disease as the vaccine patients, so neither side got a head start. Where the comparison can err, it errs against the drug. 🔬 𝗧𝗵𝗲 𝗺𝗼𝗿𝗲 𝗰𝗮𝗿𝗲𝗳𝘂𝗹 𝗮𝗻𝗮𝗹𝘆𝘀𝗶𝘀 𝗺𝗮𝗱𝗲 𝘁𝗵𝗲 𝗯𝗲𝗻𝗲𝗳𝗶𝘁 𝗯𝗶𝗴𝗴𝗲𝗿, 𝗻𝗼𝘁 𝘀𝗺𝗮𝗹𝗹𝗲𝗿 The first comparison used whole groups. A sharper one became possible once patient-level records from three other trials could be obtained, pairing each vaccine patient with controls matched on the traits that drive survival in glioblastoma, above all MGMT methylation status, matched exactly, plus age, sex, extent of surgery, residual disease, and performance status. When the comparison got sharper, the benefit grew in every one of these analyses. The original cohort-level estimate was 2.8 months. Patient-level matching across the three trials put the gain between 3.4 and 6.3 months, and a method that weights patients rather than pairing them put it between 3.4 and 4.3, with several of the matched comparisons roughly doubling the original figure. The hazard ratio moved the same way, from 0.80 to between 0.69 and 0.77. The direction matters. A real effect blurred by crude matching gets clearer when the matching improves, while a biased one tends to shrink. It got stronger. 🕵️ 𝗛𝗼𝘄 𝗺𝘂𝗰𝗵 𝗵𝗶𝗱𝗱𝗲𝗻 𝗯𝗶𝗮𝘀 𝘄𝗼𝘂𝗹𝗱 𝗶𝘁 𝘁𝗮𝗸𝗲 𝘁𝗼 𝗲𝘅𝗽𝗹𝗮𝗶𝗻 𝘁𝗵𝗶𝘀 𝗮𝘄𝗮𝘆 A fair question is how much hidden bias it would take to erase the result. Statisticians measure that with the E-value, and here a hidden factor would have to be about as strong as age is on survival, would also have to drive who received the vaccine, and would have to have escaped the decades of research that mapped every known risk factor in this disease. A second, independent check, Rosenbaum's Gamma, comes at it from the other side, asking how large an unseen imbalance between matched patients it would take to break the result, and it reaches the same verdict. Every factor strong enough to matter was already matched. A hidden one that clears that bar is not plausible. 🔒 𝗪𝗵𝘆 𝘁𝗵𝗲 𝗿𝗲𝘀𝘂𝗹𝘁 𝗶𝘀 𝗵𝗮𝗿𝗱 𝘁𝗼 𝗳𝗮𝗸𝗲 The strongest point is not any single result. It is that a hidden bias big enough to explain the effect away would have to produce the same answer in every independent test at once: • The independent randomized calibration trial. • Three separate comparison trials, drawn from different studies. • Two different statistical methods that handle the data in different ways. • A separate pooled analysis of other dendritic-cell vaccine trials. • The internal math of the survival curve, which points to the same result (about 0.71 at five years) that the patient matching found. And it would have to fall in the exact direction the biology predicted before any data existed: a slow, widening benefit concentrated in long-term survivors. A single hidden factor that could forge all of that at once is not a hidden factor. It is a coincidence that does not happen. 🩺 𝗣𝗔𝗥𝗧 𝗧𝗛𝗥𝗘𝗘: 𝗪𝗛𝗔𝗧 𝗜𝗧 𝗠𝗘𝗔𝗡𝗦 The drug is nearly free of harm, the evidence fits an established and approved regulatory path, and what remains unrun changes nothing about the case. 🛡️ 𝗧𝗵𝗲 𝘀𝗮𝗳𝗲𝘁𝘆 Two things decide whether a real effect reaches patients: whether the drug is safe enough to use, and whether regulators will accept the evidence. The first is settled. Across 2,151 doses, only five serious side effects were even possibly related to the vaccine, with no autoimmunity and no cytokine storm. It is made once, in about eight days, then stored and given as a simple injection. When a treatment barely harms, the benefit needed to justify it falls, and the benefit here clears that lower bar easily. ⏳ 𝗪𝗵𝘆 𝘁𝗵𝗲 𝘀𝘁𝗿𝗼𝗻𝗴𝗲𝗿 𝗱𝗮𝘁𝗮 𝗮𝗽𝗽𝗲𝗮𝗿𝗲𝗱 𝗼𝗻𝗹𝘆 𝗻𝗼𝘄 A reasonable person asks why the sharper analysis appeared in 2026 and not in 2023. The answer is access, not choice. The patient-level analysis was written into the trial's plan from the start, to run if and when the data could be obtained. The company tried and could not get it in 2023, because the trials that held it had not released it; it became available later through a data-sharing repository, on the data owners' timeline, not the company's. The stronger analysis was always the plan. It was waiting on data that other parties control. 🔄 𝗣𝗮𝘁𝗶𝗲𝗻𝘁𝘀 𝘄𝗵𝗼𝘀𝗲 𝗰𝗮𝗻𝗰𝗲𝗿 𝗿𝗲𝘁𝘂𝗿𝗻𝗲𝗱 𝗮𝗹𝘀𝗼 𝗯𝗲𝗻𝗲𝗳𝗶𝘁𝗲𝗱 The vaccine helped not only newly diagnosed patients but also those whose tumor had already come back, and there the effect was the largest seen anywhere in the trial. In that group, median survival ran 13.2 months from recurrence against 7.8 for the controls. The separation opened immediately, with 90.6% of vaccine patients alive at six months against 64.0% of controls, and the lead held to the later marks, where survival more than doubled: 20.7% against 9.6% at two years, 11.1% against 5.1% at two and a half. These are the original cohort-level results, already published. The high-resolution patient-level matching that was applied to the newly diagnosed group has not yet been done here, for a practical reason: it needs patient records from other recurrent-cancer trials, held by a European research organization that shares them through its own formal request. Obtaining them is a routine next step, not an obstacle, and the newly diagnosed experience suggests the sharper analysis would only make the recurrent result stronger. The approval case rests on the newly diagnosed evidence, so nothing important depends on this step; it would simply sharpen a result that is already the strongest in the trial. 🏛️ 𝗪𝗵𝗮𝘁 𝘁𝗵𝗶𝘀 𝗺𝗲𝗮𝗻𝘀 𝗳𝗼𝗿 𝗮𝗽𝗽𝗿𝗼𝘃𝗮𝗹 Regulators do not treat an external control as a first choice, but the MHRA's guideline allows it in exactly this situation: a severe disease where a placebo trial is not ethical or feasible, and an effect large enough to interpret despite the design. The guideline even gives its own worked example of an acceptable external control, and it reads almost like a description of this trial: a rare disease, no ethical placebo, same-era standard-of-care controls, an objective survival endpoint, and an effect too large to blame on bias. DCVax-L fits on every count, and it was the first medicine ever to receive the MHRA's Promising Innovative Medicine designation, which asks essentially the same questions. The newly diagnosed case carries the decision on its own evidence, and regulators weigh that evidence against the disease it treats: in a cancer this lethal, with nothing better on offer, the question is whether the benefit is large, clear, and consistent enough to act on, and a benefit of this size, pointing the same way from every direction, is. 📜 𝗧𝗵𝗶𝘀 𝗵𝗮𝘀 𝗯𝗲𝗲𝗻 𝗱𝗼𝗻𝗲 𝗯𝗲𝗳𝗼𝗿𝗲, 𝗮𝗻𝗱 𝗮𝗽𝗽𝗿𝗼𝘃𝗲𝗱 External controls are not a novelty invented for this drug. Over the past two decades they have factored into roughly forty-five drug approvals by the United States regulator, each granted under the conditions that apply here: a serious or rare disease, a placebo that would be unethical, and high unmet need. Regulators do not grant this lightly. They grant it when the disease is serious, its course is predictable and objectively measured, and the effect is large. Glioblastoma meets all three. The named cases cover every part of this disease's profile. Defibrotide, for a life-threatening transplant complication, was approved on a propensity-score comparison to a historical control, the same kind of method used here. Blinatumomab, for an aggressive relapsed leukemia, is the precedent for a fast-killing cancer, cleared by both the United States and European regulators. Cerliponase alfa is the precedent for a fatal brain disease, cleared by both agencies on treated patients versus a matched natural-history group. Glioblastoma is both at once, a fast-killing cancer of the brain, read by the same method, so no part of its profile lacks a close approved precedent. And on the one axis that governs how far an external-control result can be trusted, DCVax-L goes beyond all three. Each of those drugs was tested in a single-arm trial, with no randomization at all. DCVax-L began as a randomized trial and became an external-control comparison only when ethics consumed its placebo group. It sits in that tradition, and at the top of it. 🎯 𝗧𝗵𝗲 𝗯𝗼𝘁𝘁𝗼𝗺 𝗹𝗶𝗻𝗲 This began as a randomized trial. Medical ethics forced it into an external-control comparison. Every independent way of checking it, on different data and different math, points the same direction, and the biology predicted that direction in advance. The first analysis did not overstate the vaccine's effect. Read with the right tools, it understated it.
Andrew Caravello, DO11,406 просмотров • 3 месяцев назад
Next brain MRI scan this Friday looking 4 recurrent #glioblastoma &/or treatment effects. I don’t think high risk of major change yet but as scan creeps up closer I’m a little anxious. I understand this common in cancer patients (“scanxiety”). Here’s my take & experience to date!
Professor Richard Scolyer AO62,589 просмотров • 2 лет назад
Here’s an update on the treatment journey 4 my #glioblastoma (unfortunately is in poorest prognosis subgroup & has many extra poor prognosis molecular features)! I’m overwhelmed & v thankful for incredible support I’m receiving both medically & across the community! Sydney Local Health District
Professor Richard Scolyer AO67,394 просмотров • 3 лет назад
Best wishes 4 fab Festive Season. It’s different & more emotional 4 me than ever b4 but still enjoying it. Fingers crossed with world 1st #glioblastoma treatments I’ll b here next year & treatments/outcomes improve 4 1st time in 18 years! That would b gift of a lifetime!
Professor Richard Scolyer AO40,567 просмотров • 2 лет назад