MIRACLE-2
ForMSS rectal ca ≤10cm from anal verge, synchronous unresectable mets, 1L
ETS 76.0%
95% CI 62.4%-86.8%; single-arm, no comparator
TL;DRORR 68%, mOS 23.2mo with upfront HFRT/SBRT then chemo + tislelizumab in MSS unresectable metastatic rectal ca; 18% reached NED.
The RT-relevant claim rests on sequencing: HFRT to primary AND HFRT/SBRT to metastases delivered BEFORE any systemic therapy, with 18% (9/50) converting to NED. Dose and fractionation are not reported in source, which blocks transfer to practice. G3/4 lymphopenia 36.7% is the cost of irradiating primary plus mets ahead of cytotoxics.
In 1L MSS rectal cancer with synchronous unresectable liver or lung mets, this supports enrolling on a conversion-intent RT-first protocol rather than changing off-trial practice; it does not extend to resectable disease, MSI-high tumors, or non-rectal colon primaries.
The sequencing is the intervention: HFRT to a primary ≤10cm from the anal verge plus HFRT/SBRT to metastases delivered before any systemic therapy, with 9/50 converting to NED. Dose and fractionation are not reported in source, which blocks transfer. G3/4 lymphopenia 36.7% quantifies the cost of that combined pelvic and visceral burden.
Backbone was biomarker-gated (FOLFOX-bev for RAS/BRAF-mutant at 56.0%, FOLFIRI-cetuximab for WT) with tislelizumab 200mg Q2W added to both, so the PD-1 contribution is unmeasurable. DOR 8.0mo and 1-year DOR 20% show no durable-responder tail, arguing against real checkpoint activity in MSS disease.
Conversion is the surgical read: 18% (9/50) reached NED via primary resection plus metastasectomy or local ablation after RT-first induction, and watch-and-wait was considered for cCR where sphincter preservation failed. The 9/50 rate sets the yield to weigh when accepting these pts for staged resection planning.
| Outcome | n (%) | 95% CI |
|---|---|---|
| CR | 1 (2.0%) | n/a |
| PR | 33 (66.0%) | n/a |
| ORR | 34 (68.0%) | 53.6%-80.0% |
| DCR | 44 (88.0%) | 76.0%-95.2% |
| ETS | 38 (76.0%) | 62.4%-86.8% |
10 details 3 trials watching
Prospective single-arm phase I study at Fudan University Shanghai Cancer Center with a Fujian Cancer Hospital branch. Efficacy data available for N=50 as of Dec 31, 2025, median follow-up 19.9 months (95% CI 16.4-23.4). Reassessment every 8 weeks.
MSS rectal cancer with primary ≤10 cm from anal verge on MRI and synchronous unresectable metastases. 38 males (76.0%), median age 57 (range 30-73). Liver mets 52.0%, lung 8.0%, both 40.0%; RAS/BRAF mutant in 56.0%.
Hypofractionated RT to the primary and HFRT or SBRT to metastases, delivered first, before systemic therapy. Dose, fractionation and target volumes are not reported in source, so the RT prescription cannot be reproduced from this abstract.
After RT: FOLFOX-bevacizumab-tislelizumab for RAS/BRAF-mutant pts, FOLFIRI-cetuximab-tislelizumab for wild-type. Tislelizumab 200mg Q2W. Median eight treatment courses (range 3-12). Resectable disease went to primary resection plus metastasectomy or local ablation; watch-and-wait was considered for cCR where sphincter preservation was not possible.
Primary: ETS rate (≥20% target lesion reduction at 8 weeks post systemic initiation). Secondary: DCR, DOR, OS, PFS, safety. Note the headline ORR and OS circulating with this abstract are secondary, not the registered primary.
No grade 5 TRAEs. All-grade lymphopenia 95.9%, anemia 91.8%, leukopenia 69.4%. G3/4: lymphopenia 36.7%, neutropenia 26.5%, leukopenia 20.4%. The marrow and lymphoid toxicity dominates, which is the expected signature of irradiating a pelvic primary plus liver/lung mets ahead of FOLFOX or FOLFIRI.
The RT prescription is absent from the abstract, so the intervention cannot be replicated. DOR median 8.0 months with a 1-year DOR rate of 20% sits awkwardly under a 68% ORR, and PFS 9.3 months against OS 23.2 months means most of the survival curve is post-progression, where subsequent lines rather than the studied regimen are acting.
The trial's own conclusion claims only a high ETS rate and manageable safety, not an immune-resistance mechanism. Whether RT contributed anything beyond debulking is untestable in a single-arm design where every patient also received chemotherapy plus a PD-1 antibody.
| Endpoint | Median | 95% CI | 1-yr rate |
|---|---|---|---|
| OS | 23.2 mo | 15.1-31.3 | 93.3% |
| PFS | 9.3 mo | 7.1-11.5 | 33.4% |
| DOR (n=34 CR/PR) | 8.0 mo | 5.2-10.8 | 20% |
Single-arm phase I, N=50, no randomised comparator; median f/u 19.9mo with OS still immature. Hard maturity gate: single-arm never reaches confirmatory.
- Does RT add anything over chemo plus PD-1 alone in MSS mCRC? recruiting Regorafenib Alone or in Combination With Hypofractionated/Low-dose Radiotherapy Plus Toripalimab for Metastatic Colorectal Cancer Phase 2n=108 · primary completion 2025-04 · randomised: regorafenib +/- HFRT/LDRT + toripalimabnot yet A Combination Therapy Including Anti-PD-1 Immunotherapy in MSS Rectal Cancer With Resectable Distal Metastasis Phase 2n=52 · primary completion 2027-09 · SCRT then tislelizumab combo, MSS rectal w/ mets
- Durability beyond 1 year given 20% 1-yr DOR rate
- Optimal RT dose and fractionation for primary plus metastases recruiting Standard Systemic Therapy Combined With High/Low-dose Radiotherapy Plus Toripalimab for Metastatic Colorectal Cancer Phase 2n=96 · primary completion 2026-12 · TORCH-M: high vs low-dose RT + toripalimab, MSS mCRC
📚 Sources · 🐦 1 tweet
MIRACLE-2: RT to primary/mets -> chemo + tislelizumab in MSS unresectable met rectal ca (N=50): 68% ORR & median OS 23 mo.
— Dr. Nina Niu Sanford (@NiuSanford) May 31, 2026
Early, single-arm data, but ~1 in 5 pts reached NED.
Suggests RT + systemic + PD1 blockade could overcome immune resistance in MSS mCRC. #ASCO26 @OncoAlert pic.twitter.com/sjnUW8x7f3
The longer read
The claim being made around this abstract, that radiotherapy overcomes immune resistance in microsatellite-stable colorectal cancer, is not a claim this trial design can support, and separating the interesting part from the overreach is most of the work of reading it. Every one of the 50 patients received radiotherapy, then chemotherapy, then tislelizumab. There is no arm in which any component is withheld. An ORR of 68% in first-line metastatic colorectal cancer is broadly in the range chemotherapy plus a biologic already delivers in this setting, so the response rate alone carries no information about what the PD-1 antibody or the radiation added. The abstract's own conclusion is more careful than the surrounding commentary: it claims a high ETS rate and manageable safety, nothing about immune priming.
What is genuinely unusual here is the sequencing and the conversion intent. Radiotherapy went to the primary and to the metastases up front, before any systemic therapy, in a population defined by unresectable synchronous disease. The conventional order in this setting is systemic therapy first, with local treatment reserved for responders. Putting hypofractionated radiotherapy to a low rectal primary and stereotactic or hypofractionated treatment to liver and lung metastases ahead of FOLFOX or FOLFIRI is a real departure, and 9 of 50 patients reaching no evidence of disease is the number that makes the strategy worth a randomised test. That is a conversion result, not an immunological one, and it should be framed that way.
The durability data argue for restraint. Median duration of response was 8.0 months among the 34 responders, with a 1-year duration-of-response rate of 20%, and median PFS was 9.3 months. Set against a median OS of 23.2 months, most of the survival time accrues after progression, which is where later lines of therapy operate rather than the studied regimen. A checkpoint inhibitor that was meaningfully converting an immunologically cold tumor would be expected to produce a tail of durable responders; a 20% 1-year duration-of-response rate is not that shape. The more parsimonious read is deep, chemotherapy-driven cytoreduction augmented by local ablation of every visible site, which is a legitimate and useful thing to achieve, just not the mechanism being advertised.
For a radiation oncologist the central problem is that the abstract does not state dose, fractionation, or target volume for either the primary or the metastatic sites. That omission makes the intervention unreproducible in the specific way that matters most: hypofractionated pelvic radiotherapy delivered before FOLFOX is a different toxicity proposition depending on whether it was a short course or an extended one, and grade 3/4 lymphopenia of 36.7% is a signal that the combined pelvic and visceral irradiation burden was not trivial. Lymphopenia of that magnitude is also mechanistically awkward for the immune-priming hypothesis, since the effector population being counted on is the one being depleted.
What would have to be true for this to be more than a promising single-arm series is a randomised comparison of radiotherapy-first against systemic-first with the same conversion intent, powered on a conversion or NED endpoint rather than response rate. The study's own stated future direction, predictive biomarker discovery, is the right instinct but the wrong priority order: the sequencing question comes first, because without it there is no established effect for a biomarker to select for.