This Week in Hematology — Sep 23, 2026

Welcome to This Week in Hematology. This week we're covering 10 notable papers spanning acute leukaemia therapy and risk stratification, maintenance strategies and prognostic modelling in myeloma and CLL, and a pair of practical updates in thrombosis, plus a first-in-human mRNA immunotherapy. Let's dive in.

We start with acute myeloid leukaemia, where a rare subtype may be about to change categories. In Blood, Bertoli and colleagues report a retrospective series of 212 adults with de novo BCR::ABL1-positive AML, treated between 1999 and 2024 — a group currently classed as adverse risk. Eighty-nine received intensive chemotherapy alone, 123 received chemotherapy plus a tyrosine kinase inhibitor. Adding the TKI raised complete remission, with or without count recovery, to about 90 percent, compared with roughly two thirds on chemotherapy alone. Median overall survival was 20.5 months with chemotherapy alone and had not been reached in the TKI group, with five-year survival of about 63 percent versus around 39 percent. On multivariable analysis the TKI independently roughly quintupled the odds of remission and cut the risk of death by about 60 percent, and both the TKI and allogeneic transplant in first remission independently improved relapse-free survival. This is registry-style data, not a randomised trial, but the effect size is large and consistent, and the authors argue that TKI plus intensive chemotherapy should be standard of care and that the adverse-risk label needs revisiting in the next European LeukemiaNet classification.

Staying with AML but at the other end of the fitness spectrum, the American Journal of Hematology published a Mayo Clinic analysis by Kumar and colleagues of 540 newly diagnosed patients treated with venetoclax plus a hypomethylating agent, comparing venetoclax durations in cycle one of seven, fourteen, twenty-one, or twenty-eight days. With a median follow-up of just over three years, transplant-censored survival did not differ significantly across the four schedules, and transplant rates were similar. What drove outcome was genetics — survival ranged from about six months in high-risk disease to over eighteen months in favourable-risk by European LeukemiaNet 2024 criteria, and the Mayo genetic model separated patients just as sharply. One caveat worth carrying into clinic: among high-risk patients, the fourteen-day schedule looked inferior to twenty-one and twenty-eight days, and early mortality at thirty and sixty days was actually higher with the shorter schedules, which the authors attribute to selection bias — the frailest patients got the shortest courses. The practical message is that duration shortening for tolerability does not appear to cost survival in most risk groups, but this is retrospective and prospective risk-adapted dosing trials are still needed. Complementing both papers, Hourigan and co-authors offer a How I Treat piece in Blood on using measurable residual disease to personalise allogeneic transplant in adult AML — patient selection, conditioning intensity, post-transplant surveillance and pre-emptive therapy, worked through four cases. Their tone is appropriately sober: MRD is useful in specific contexts, but they urge separating translational aspiration from what assays can actually deliver reproducibly today.

Next, two randomised maintenance trials, both in the British Journal of Haematology, that land in opposite directions. The CLL6 RESIDUUM study, reported by Aurran-Schleinitz and colleagues from the Australasian and French cooperative groups, randomised 143 patients with chronic lymphocytic leukaemia who still had measurable residual disease after frontline immunochemotherapy to two years of daily lenalidomide or observation. Median progression-free survival was about 65 months with lenalidomide versus roughly 42 months with observation, a statistically significant gain that persisted for three years after maintenance stopped, and more patients on lenalidomide converted to MRD negativity, particularly those completing at least twenty cycles. There was no overall survival difference, and lenalidomide brought more cytopenias, infections and gastrointestinal toxicity. Note also the era — this is an immunochemotherapy-treated population, so the place of this strategy alongside modern targeted frontline therapy is an open question. Contrast that with SeaLAND, the phase three ALLG MM23 trial led by Rees, which added low-dose weekly selinexor to lenalidomide maintenance after autologous transplant in newly diagnosed myeloma. The trial closed early for futility. Among 142 patients, complete response rates were numerically higher with the combination but not significantly so, and progression-free survival at two years was essentially identical between arms. Meanwhile grade three or worse adverse events were roughly twice as common with selinexor, with more severe infections and gastrointestinal toxicity, and crucially the dose intensity of lenalidomide itself fell from about 81 percent to 68 percent when selinexor was added. The authors conclude selinexor-lenalidomide maintenance cannot be recommended, including in high-risk disease. The lesson generalises: a maintenance partner that erodes delivery of the backbone drug is unlikely to help.

Two further myeloma papers push on risk stratification. In Blood, Toho and colleagues studied 352 newly diagnosed patients, combining bone marrow plasma cell percentage from whole slide images, diffusion volume on MRI, metabolic tumour volume on PET, and serum soluble BCMA. Using unsupervised clustering they defined four phenotypes along two axes — physical tumour volume and soluble BCMA. Soluble BCMA was the strongest and most consistent prognostic marker, independent of imaging burden, and it, rather than tumour volume, predicted early treatment failure: patients with low imaging volume but high soluble BCMA had twelve-month progression-free survival similar to those with high burden on both axes. They also describe a spatially sequestered phenotype with extensive imaging disease but low soluble BCMA, in which gain of 1q21 lost its prognostic weight — whereas in the serologically dominant group it strongly predicted worse survival. Alongside that, Zeng and colleagues in the British Journal of Haematology analysed 694 newly diagnosed Chinese patients with SNP-array and FISH, and found hyperdiploidy is far from uniform. A modal chromosome count above 49 predicted better overall and progression-free survival, as did gain of chromosome 3 — and importantly, hyperdiploidy rescued prognosis in patients with isolated 1q21 gain but not in those carrying t(4;14), t(14;16), 17p deletion or 1p32 loss. Autologous transplant improved survival only in the low-risk, higher-ploidy hyperdiploid group. Their pragmatic suggestion is that FISH for trisomy 3 could serve as a cheap, fast surrogate for ploidy assessment.

On the thrombosis side, the Journal of Thrombosis and Haemostasis carries two clinically oriented reviews. Spyropoulos and Douketis address perioperative management of patients on factor XI and factor XIa inhibitors — antisense oligonucleotides, monoclonal antibodies and small molecules that attenuate thrombosis while largely sparing haemostasis. Drawing on the surgical phenotype of congenital factor XI deficiency and completed phase two data, they argue these agents may allow venous thromboprophylaxis in major orthopaedic surgery with both less thromboembolism and less clinically relevant bleeding than enoxaparin, and raise the prospect of not interrupting therapy at all for low and some moderate bleeding-risk procedures. That remains a hypothesis: dedicated perioperative management studies, reversal strategies and usable assays are all still missing. Separately, Wang, Le Gal and Carrier tackle subsegmental pulmonary embolism, which makes up something like five to ten percent of PE diagnoses and where the decision to anticoagulate remains genuinely unsettled — they work through cases, give evidence-based recommendations where they exist, and flag that a randomised trial of anticoagulation in isolated subsegmental PE is ongoing.

Finally, a striking early-phase result in Cell. Wang and colleagues describe ABO2203, a lipid nanoparticle-delivered messenger RNA encoding a CD19-targeting T cell engager. In transgenic mice it produced complete B cell depletion with less cytokine release than the engager given as protein, and in a first-in-human study of just three patients with refractory secondary immune thrombocytopenia it achieved rapid, complete peripheral B cell depletion with sustained marrow depletion. All three patients had durable platelet recovery and improved serology through six months, with only grade one and two adverse events and no cytokine release syndrome. Three patients is three patients — but this is proof of concept that an mRNA-encoded engager can do the work of CAR-T or protein engagers in autoimmune cytopenias, with a potentially gentler safety profile.

If you only have time for one paper this week, make it the Blood analysis of tyrosine kinase inhibitors added to intensive chemotherapy in de novo BCR::ABL1-positive AML. It is the one paper here that should change what you prescribe on Monday for a specific, identifiable patient — and it may well change how this subtype is risk-classified.

Here are the key takeaways from this week in Hematology. In de novo BCR::ABL1-positive AML, add a TKI to intensive chemotherapy — remission rates and survival are substantially better, and transplant in first remission still adds benefit. With venetoclax plus a hypomethylating agent, genetics drives outcome far more than venetoclax duration, though be cautious about truncated schedules in high-risk disease. In myeloma maintenance, selinexor added to lenalidomide failed on progression-free survival and doubled severe toxicity while reducing lenalidomide dose intensity — don't use it. In CLL with residual disease after immunochemotherapy, lenalidomide maintenance prolongs progression-free survival and deepens response, but with no survival gain and real toxicity. And for myeloma risk assessment, soluble BCMA adds information that imaging burden does not, while among hyperdiploid patients chromosome count above 49 and trisomy 3 mark a genuinely better-prognosis group.

That's your roundup for This Week in Hematology. The full transcript and references are available on the episode page in your AudioScholar library. This is an AI-curated summary — for clinical decisions, always consult primary sources and current guidelines. See you next week.

If this weekly briefing is useful, follow the show in your podcast app so new episodes arrive automatically. And think of one colleague — in any specialty — who never has time to keep up with the literature. Tell them about AudioScholar: a free ten-minute weekly for every specialty, to listen to in any podcast app, or to read at audioscholar dot C C.

This is an automated summary generated by artificial intelligence, which can make mistakes. Always review the original source materials.