Sascha Roth and the Trial Where Every Tumor Vanished (2022)
The story
Skip to the verdict ↓Sascha Roth was packing on a Friday evening, racing to reach New York for weeks of radiation on her rectal cancer, when her doctor called and told her to stop — a fresh round of scans had come back, and a team at Memorial Sloan Kettering wanted to see her before she went any further. Roth, thirty-eight, had been the first patient enrolled in a small trial of a drug called dostarlimab.
What that team — Memorial Sloan Kettering Cancer Center's — had seen in the scans was this: they could not find any cancer. There was no reason to put her through radiation.
Roth helps run a family home-furnishings business in the Bethesda, Maryland area, and the trial she had joined was of an immunotherapy. When her own result held, the researchers worried she might be a fluke. She was not.
The Trial
The trial, led by Dr. Andrea Cercek and Dr. Luis Diaz Jr., enrolled patients with a specific subtype of rectal cancer: mismatch repair-deficient (sometimes written MMRd, or microsatellite-instability-high), stage II or III. That subtype is about 5 to 10 percent of rectal cancers and tends to do badly on chemotherapy. Patients got dostarlimab, an anti-PD-1 antibody, intravenously every three weeks for six months. The plan was that anyone whose cancer did not respond would move on to the standard treatment — surgery, radiation, chemotherapy.
In the results published in the New England Journal of Medicine on June 23, 2022, all 12 patients who had finished treatment with at least six months of follow-up showed a clinical complete response. No tumor was detectable on physical exam, endoscopy, PET, or MRI. None of them needed surgery, radiation, or chemotherapy. There were no severe side effects. Diaz told the New York Times, 'I believe this is the first time this has happened in the history of cancer.'
The field's own caution came in the same week. Dr. Hanna Sanoff of the University of North Carolina, writing separately, noted that a clinical complete response 'is not a surrogate' for a long-term cure, and that the cohort was small and the follow-up short.
The Mechanism
Checkpoint-inhibitor drugs take the brakes off the immune system, letting it see and attack tumors it had been ignoring. Mismatch repair-deficient tumors carry a very high number of mutations, which makes them unusually visible to that unleashed immune response. The cohort was selected precisely because their tumors were the kind most likely to respond. The program has since expanded into larger trials, and the drug drew an FDA breakthrough-therapy designation in December 2024.
It happened — and nature accounts for it.
Reviewer Notes
Miracles Jar weighs each claim two ways — how extraordinary it would be, and how strong the evidence is.
Assessed by Miracles Jar AI
Astonishing as medicine, unremarkable as metaphysics: a peer-reviewed immunotherapy trial in which all 12 patients with a specific rectal-cancer subtype reached complete remission through a fully understood mechanism. Scored near the floor; the case shows what a fully explained mass remission looks like.
Mode and score. This entry asks whether nature could explain it — and here, it plainly can. The case is scored near the floor because the result is explained: a known drug produced a known kind of effect in a selected, genetically defined cohort. The case is presented deliberately, inverted from the catalog's other healing cases: most of the catalog's healing entries are claims that would have no natural explanation if the facts held, and this case is the inverse, which is what makes it the right comparison. The facts are not in doubt; they are peer-reviewed. And the explanation is not missing; it is known. A natural explanation is almost certain.
The weighing. A 100 percent response is a stunning effect size. It is still a drug doing a thing drugs are understood to do, in a cohort selected precisely because their tumors were the kind most likely to respond. There is no suspension of natural law here and no improbable timing, only a mechanism that worked. That is not a verdict on the result, which is one of the more remarkable things modern oncology has produced — it is a verdict on the category.
Why it belongs in the catalog. The case shows what a fully documented, mechanistically understood mass remission looks like, so that prayer-associated remissions can be measured against a real benchmark rather than against intuition. When a remission that followed prayer is offered as a miracle, this is the comparison to make first — a documented drug, a named cohort, an understood mechanism, peer review, and an effect medicine can explain and set out to repeat. The honest question for any unexplained remission is how far short of this standard of explanation it falls, and why.
Evidence summary. The result is peer-reviewed in NEJM — all 12 patients with at least six months of follow-up reached a clinical complete response, no tumor detectable on exam, endoscopy, PET, or MRI; the facts are not in dispute and the documentation sits at the top of the medical evidence hierarchy (neutral, strong). The mechanism is known and published — checkpoint-inhibitor immunotherapy unmasks tumors to the immune system, and MMRd tumors are unusually visible because of their high mutational load; an extraordinary effect size but a pharmacological effect with an understood basis, not an event without a natural account (natural, strong). The effect is reproducible in principle and being tested in larger cohorts, with an FDA breakthrough-therapy designation in December 2024 as the program expanded; repeatability separates a drug effect from a one-off anomaly (natural, moderate). Independent caution is on the record — Dr. Hanna Sanoff noted that a clinical complete response is not a surrogate for long-term cure and that longer follow-up was needed; even the celebrated result was framed by its own field as preliminary, the opposite of a miracle claim's usual posture (neutral, moderate). There is no improbable coincidence and no suspension of natural law — a known drug produced a known kind of effect in a selected, genetically defined cohort; it shows what a fully explained mass remission looks like, against which prayer-associated remissions can be compared (natural, strong).
The trial was run by Memorial Sloan Kettering Cancer Center. Dostarlimab's brand name is Jemperli, an anti-PD-1 monoclonal antibody made by GSK. The complete-response rate carried a 95 percent confidence interval of 74 to 100 percent, with no grade 3 or higher adverse events, over a follow-up that then ranged from 6 to 25 months. The NEJM citation is Cercek et al., PMID 35660797. Cost was roughly $11,000 per dose. Lead researcher Diaz called the result 'practice-changing,' and Cercek spoke about the permanent effects of surgery and radiation that patients would now avoid. Roth remained cancer-free for years afterward. Diaz is the source of the 'first time this has happened in the history of cancer' quotation.
Evidence ledger — what the verdict rests on
The result is peer-reviewed in the New England Journal of Medicine: all 12 patients with at least six months of follow-up reached a clinical complete response, with no tumor detectable on exam, endoscopy, PET, or MRI
The facts are not in dispute; the documentation is at the top of the medical evidence hierarchy
The mechanism is known and published: checkpoint-inhibitor immunotherapy unmasks tumors to the immune system, and mismatch repair-deficient tumors are unusually visible because of their high mutational load
The 100 percent response is an extraordinary effect size, but it is a pharmacological effect with an understood basis, not an event without a natural account
The effect is reproducible in principle and is being tested in larger cohorts; the drug received an FDA breakthrough-therapy designation in December 2024 as the program expanded
Repeatability is the property that separates a drug effect from a one-off anomaly
Independent caution is on the record: Dr. Hanna Sanoff noted that a clinical complete response is not a surrogate for long-term cure and that longer follow-up was needed
Even the celebrated result was framed by its own field as preliminary, which is the opposite of a miracle claim's usual posture
There is no improbable coincidence and no suspension of natural law: a known drug produced a known kind of effect in a selected, genetically defined cohort
It shows what a fully explained mass remission looks like, against which prayer-associated remissions can be compared
What would raise the meter: Long-term follow-up documenting permanence, in a condition with a near-zero spontaneous-resolution base rate, would raise the meter.
What would lower it: A documented relapse, or case literature showing the condition fluctuates or remits on its own, would move it down.
How we weigh every claim — the full method, deductions and all →
The natural explanation
The leading natural account for this case is spontaneous remission & the body's own recovery. Read what it explains — and where it stops.
Sources
Tagged by proximity to the event. Primary sources are direct or contemporaneous; tertiary are downstream retellings.
- 1.Primaryacademic
The peer-reviewed source: 12 patients, 100 percent clinical complete response (95% CI, 74–100), MMRd stage II/III rectal adenocarcinoma, dostarlimab IV every 3 weeks for 6 months, follow-up 6–25 months, no grade 3+ adverse events
- 2.Primarywebsite
Investigators Cercek and Diaz, the MMRd/MSI subtype at roughly 5–10 percent of rectal cancers, Sascha Roth as first patient at age 38 with a family home-furnishings business and years of cancer-free follow-up, and the December 2024 FDA breakthrough-therapy designation
- 3.Secondarynews
The undetectable result across exam, endoscopy, PET and MRI; roughly $11,000 per dose; Cercek on the permanent effects of surgery and radiation; Diaz calling it 'practice-changing'; and Sanoff's caution that longer follow-up is needed
- 4.Secondarynews
Sascha Roth as the first enrollee and Cercek's call that radiation was unnecessary; Diaz's 'first time this has happened in the history of cancer'; and Sanoff's caveat that a clinical complete response 'is not a surrogate' for long-term control
Cases like this
Nearest on the map — similar in how miraculous they’d be, and how strong the evidence is.